Bacopa Monnieri (Brahmi)

Overview

Bacopa monnieri is a creeping wetland herb from Ayurvedic tradition whose active saponins, the bacosides, have been tested in at least nine placebo-controlled trials for memory in healthy adults. At 300 to 320 mg a day of a standardized extract, the consistent finding is better retention of newly learned information after about 12 weeks. The catch: it does nothing in an afternoon, and the most common complaint in the trials was an unhappy stomach.

What is Bacopa Monnieri (Brahmi)?

Key takeaways

• Every 12-week placebo-controlled trial that measured delayed word recall found more of it on 300 to 320 mg a day of a standardized extract: Stough 2001, Roodenrys 2002 (n=76), Calabrese 2008 (n=54), and Morgan 2010 (n=98).
• Bacopa is not a stimulant. A single 300 mg dose did nothing measurable two hours later (Nathan 2001), and the one trial that tested both time points found the memory effect larger at 12 weeks than at 5 (Stough 2001).
• The pooled effect on speed of attention is real and small: 17.9 milliseconds faster on a trail-making test and 10.6 milliseconds faster on choice reaction time across nine trials and 518 people (Kongkeaw 2014, J Ethnopharmacol).
• Synapsa, KeenMind, and CDRI 08 are the same extract under three names, standardized to 55% bacosides and used at 300 to 320 mg in the Australian trials. A generic "50% bacosides" label is usually a UV-spectrophotometry number that counts molecules that are not bacosides.
• The most common side effect in every trial that reported one is digestive: more frequent stools, abdominal cramps, and nausea (Morgan 2010), stomach upset (Calabrese 2008), mild gut complaints that resolved on their own (Pravina 2007). Taking it with a meal is the standard fix.
• Bacopa raised thyroxine 41% in mice at 200 mg/kg (Kar 2002). Nobody has tested that in people, which is exactly why anyone on thyroid medication should tell their prescriber before starting it.

Bacopa monnieri is a small, fleshy-leaved herb that creeps along the margins of ponds, rice paddies, and slow streams across India, Southeast Asia, northern Australia, and the warm coasts of the Americas. In Ayurveda it is Brahmi, named for Brahma, and it belongs to the medhya rasayana, the class of plants prescribed for intellect and memory; it appears in the Charaka Samhita, which makes the claim at least a millennium old. The modern claim is narrower and more testable: a standardized extract, taken daily for about three months, helps healthy adults hold on to what they have just learned. Aquarium shops sell the same plant as moneywort, which is the least mystical thing you will read about it today.

The molecules doing the work are the bacosides. A bacoside is a triterpenoid saponin: a large, soap-like molecule built on a 30-carbon backbone (the triterpene) with one or more sugar chains hung off it (the glycoside). The plant makes dozens of them, with names like bacoside A3, bacopaside I, bacopaside II, and bacopasaponin C. "Bacoside A" on a label is not one compound but a group of at least four, and "bacosides A and B" is the older, cruder way of naming the whole family. Bacopa also carries small amounts of flavonoids such as luteolin and quercetin, which turn up in the mechanism papers but at amounts too small to matter on their own.

Here is the surprise: the number on a bacopa label depends more on the lab method than on the plant. Bacosides absorb ultraviolet light weakly, so the old UV spectrophotometry assay cannot tell a bacoside from any other saponin, degradation product, or plant residue that absorbs at the same wavelength. The number it produces is a "total saponins" figure wearing a bacoside costume. High-performance liquid chromatography (HPLC), the method the United States and British Pharmacopoeias specify, separates and counts named marker compounds, and an extract that reads 50% by UV can read a small fraction of that by HPLC. Our full review of the Synapsa extract walks through both methods; the short version is that "50% bacosides" and "55% bacosides" can describe two extracts that have almost nothing in common.

Where the extract comes from matters for the same reason. The whole aerial plant is harvested, dried, extracted with water, alcohol, or a blend, and concentrated. CDRI 08 is the extract developed at India's Central Drug Research Institute; it is the name on the Swinburne University trial papers under its Australian trade name, KeenMind, and it is licensed in the United States as Synapsa. BacoMind, from Natural Remedies, is the extract in the largest single trial (Morgan 2010). Bacognize is a third standardized extract with one published trial (Kumar 2016). Those three names, plus "generic," cover the whole forms table below, and the difference between them is not marketing. It is which trial you are allowed to point to.

How the body handles bacosides is the honest gap in this page. No published human pharmacokinetic study of bacosides turned up in a PubMed search on the day this page was built, so nobody can tell you the peak blood level, the half-life, or the fraction that crosses into the brain. Saponins as a class are large, poorly absorbed molecules that stomach acid and gut bacteria partly break apart, clipping off the sugars to leave the backbone (the aglycone); for bacosides, those fragments are the bacogenins and the jujubogenin and pseudojujubogenin derivatives. Whether the intact bacosides or their fragments are the active species in a person has not been shown. What has been shown is the timing. Effects appear over weeks, not hours, which fits a molecule that works by slowly changing cells rather than by hitting a receptor and leaving.

One more naming note, because the search box brings people here on it: "water hyssop" is an old common name for Bacopa and has nothing to do with the kitchen herb hyssop, which is a mint relative. If a product says hyssop and means hyssop, it is a different plant with none of the trials on this page behind it.

How it works

Nobody has a complete answer, and the most-cited review says so. Aguiar and Borowski's 2013 synthesis in Rejuvenation Research lists five mechanisms with animal or cell support: antioxidant protection of neurons through enzyme induction, inhibition of acetylcholinesterase or activation of choline acetyltransferase, increased cerebral blood flow, modulation of the acetylcholine, serotonin, and dopamine systems, and reduction of amyloid protein in animal models. Five candidate mechanisms is a polite way of saying the field has not picked one. Here is what each would mean for a person, in the order the evidence supports them.

Start with acetylcholine, because it is the only mechanism with a human measurement. Acetylcholine is the neurotransmitter that attention and memory circuits run on, and acetylcholinesterase (AChE) is the enzyme that clears it from the synapse. Slow the enzyme and acetylcholine lingers longer, which is the same logic behind several prescription cognitive drugs. In the one trial that checked, 60 healthy adults averaging 62 years old who took 300 or 600 mg of a standardized extract for 12 weeks showed suppressed plasma AChE activity alongside faster brainwave responses (shorter N100 and P300 latencies, two event-related potentials that track how quickly the brain registers and categorizes a stimulus) and better working memory (Peth-Nui 2012). That is a biomarker in blood, not a measurement in brain tissue, but it is the mechanism's only human footprint.

The second mechanism is structural, and it explains the calendar. In rats, bacopa extract raises the activity of calcium-dependent kinases in the synapse and the post-synaptic cell, the enzymes that strengthen existing connections between neurons and grow new ones, particularly in the hippocampus, the region that files new memories (reviewed in McPhee 2016). Growing dendrites and synapses is slow work measured in weeks. A stimulant changes how a neuron fires in minutes; a molecule that changes how many connections a neuron has needs the time it takes to build them. That proposed mechanism is why a 12-week trial finds something a two-hour test does not, and it is still labeled proposed because no human imaging study has watched it happen.

The third is antioxidant and anti-stress, and it is the weakest link to the outcomes people buy bacopa for. Bacosides induce antioxidant enzymes in rodent brain, and one small acute study found lower salivary cortisol and better mood after 320 mg of CDRI 08 during a stressful multitasking test (Benson 2014). What would turn this from proposed into shown is a trial in which cortisol or an oxidative-stress marker is the primary outcome and the memory benefit tracks with it. That trial has not been run.

What a person might notice, then, is not a buzz. In the trials the difference showed up on tests: more words recalled from a list after a delay, fewer forgotten between sessions, reaction times a few hundredths of a second faster. Subjective memory, what people say about their own recall on a questionnaire, improved in the 98-person BacoMind trial but not by enough to reach significance (Morgan 2010). Expect the lab test to notice before you do.

Each of those mechanisms points at a different outcome, and the table below grades them separately: retention of new information for the synaptic mechanism, speed of attention for the cholinergic one, and anxiety scores and acute performance for the stress mechanism. Citicoline, the ingredient most often stacked with bacopa, works on a different step entirely (it supplies choline and the raw material for neuron membranes), which is the reason the two are paired rather than a reason to expect one to do the other's job.

Evidence at a glance

What the evidence doesn't support yet

The same-day claim is the first one to drop. "Take it before the exam" is how bacopa gets sold to students, and the best-designed single-dose trial, 38 people given 300 mg or placebo and tested two hours later, found no difference on any test (Nathan 2001). The two crossover studies that did find acute effects used 320 mg of CDRI 08 in 24 and 17 people, on some tasks and not others, with the authors of both calling for replication in a larger sample (Downey 2013, Benson 2014). Forty-one people and a split result is a lead, not a benefit.

Focus is the second. Bacopa is routinely sold for attention, and the data say memory. The systematic review that looked domain by domain found enhancement on free recall and "little evidence of enhancement in any other cognitive domain" (Pase 2012). The 2014 meta-analysis found faster reaction times, but by 10 to 18 milliseconds, which no one will feel. The 2026 network meta-analysis of 29 trials found no significant effect on sustained attention, selective attention, or processing speed at any dose (Tiemtad 2026). A label that leads with focus is leading with the weakest finding.

Third, nothing on this page supports taking bacopa for years to protect memory as you age. The longest completed adult trials ran 12 to 13 weeks in healthy volunteers. The one trial built to answer the long-term question, ARCLI, was registered in 2011 to follow 465 older adults for 12 months on 300 mg of CDRI 08, and its cognitive results had not appeared in PubMed when this page was built (Stough 2012). Until they do, "long-term brain health" is an extrapolation.

Fourth, the stress and calm claims rest on secondary outcomes. Two 12-week trials saw anxiety questionnaire scores fall (Stough 2001, Calabrese 2008), one saw no change (Roodenrys 2002), and the acute cortisol drop came from 17 people (Benson 2014). No trial has recruited stressed adults and made stress the primary endpoint. If stress rather than memory is the goal, start with what the ashwagandha trials measured instead.

Clinical evidence

• Meta-analysis of randomized controlled trials on cognitive effects of Bacopa monnieri extract — Nine randomized, placebo-controlled trials of standardized Bacopa monnieri extract taken for 12 weeks or longer without co-medication, 518 participants in total, 437 in the pooled analysis. Measured Trail B time and choice reaction time as the pooled cognitive outcomes. Bacopa shortened Trail B by 17.9 milliseconds (95% CI 24.6 to 11.2) and choice reaction time by 10.6 milliseconds (95% CI 12.1 to 9.2), both p<0.001. All trials were low risk of bias. Limitation: the pooled effects are speed-of-attention measures in the tens of milliseconds, and the authors say only a large head-to-head trial would settle efficacy.
• Comparative effects of Bacopa monnieri and Ginkgo biloba on cognitive functions: a systematic review and network meta-analysis — Twenty-nine randomized controlled trials in healthy adults, 2,107 participants, comparing bacopa and ginkgo extracts to each other and to placebo by indirect (network) comparison. Measured working memory, short-term memory, delayed memory, sustained and selective attention, and processing speed as standardized mean differences. High-dose bacopa (600 mg/day or more) beat placebo on working memory (SMD 2.03, 95% CI 1.28 to 2.78) and on short-term memory; low-dose bacopa (300 to under 600 mg) beat placebo on delayed memory; no dose of either herb changed sustained attention, selective attention, or processing speed. Limitation: the dose finding comes from indirect comparison across trials, not a dose-ranging study, and the authors note the lack of direct head-to-head trials.
• The cognitive-enhancing effects of Bacopa monnieri: a systematic review of randomized, controlled human clinical trials — Systematic review without pooling of six randomized controlled trials in adults without cognitive impairment, each 12 weeks long, using three different bacopa extracts at 300 to 450 mg per day. Catalogued every validated cognitive test by domain. Bacopa improved performance on 9 of 17 tests of memory free recall; there was little evidence of enhancement in any other domain, and no trial had tested auditory perception, idea production, or language. Limitation: six small trials with inconsistent batteries, so the absence of effects outside memory may reflect what was never measured.
• Does Bacopa monnieri improve memory performance in older persons? Results of a randomized, placebo-controlled, double-blind trial — 98 healthy adults over 55 in Lismore, Australia (81 completed), 12 weeks, 300 mg/day of BacoMind extract versus identical placebo. Measured the Rey Auditory Verbal Learning Test (AVLT), the Rey-Osterrieth Complex Figure Test, Trail Making, and a subjective memory questionnaire. Bacopa improved AVLT learning trials, total learning (p=0.011), delayed recall (p=0.001), and resistance to retroactive interference (p=0.048); figure recall, trail making, and subjective memory improved without reaching significance. Bacopa caused more gastrointestinal side effects than placebo: increased stool frequency, abdominal cramps, and nausea. Limitation: a single extract and a single 12-week window, with 17 dropouts.
• Effects of a standardized Bacopa monnieri extract on cognitive performance, anxiety, and depression in the elderly: a randomized, double-blind, placebo-controlled trial — 54 adults aged 65 and older (mean 73.5) in Portland, Oregon, without clinical signs of dementia, 48 completers; 6-week placebo run-in then 12 weeks of 300 mg/day standardized whole-plant extract or placebo. Primary outcome AVLT delayed recall; also Stroop, divided attention, a working memory digit task, anxiety, depression, and mood scales, heart rate, and blood pressure. Bacopa improved AVLT delayed recall and Stroop performance relative to placebo after adjusting for baseline; combined anxiety scores, a depression scale score, and heart rate decreased over time on bacopa and increased on placebo; no effect on divided attention, the digit task, mood, or blood pressure. Adverse events were 9 on bacopa and 10 on placebo, primarily stomach upset. Limitation: 24 per arm, and the mood findings were secondary outcomes.
• The chronic effects of an extract of Bacopa monniera (Brahmi) on cognitive function in healthy human subjects — Healthy adults in a double-blind, placebo-controlled, independent-groups design, 12 weeks, 300 mg/day of the KeenMind (CDRI 08) extract or placebo, tested at baseline, 5 weeks, and 12 weeks. Measured speed of visual information processing (inspection time task), the AVLT for learning rate and memory consolidation, and state anxiety. Bacopa improved inspection-time speed, learning rate, and memory consolidation (p<0.05) and reduced state anxiety (p<0.001) versus placebo, with maximal effects at 12 weeks. Limitation: the abstract does not report the sample size, and the effects were larger at 12 weeks than at 5, which is a single data point for the time course.
• Examining the nootropic effects of a special extract of Bacopa monniera on human cognitive functioning: 90 day double-blind placebo-controlled randomized trial — 107 healthy adults recruited, 62 completed with 80% compliance, 90 days, 2 x 150 mg/day of KeenMind (CDRI 08) or placebo. Measured the Cognitive Drug Research assessment system at baseline and day 90. Bacopa improved the Working Memory factor, specifically spatial working memory accuracy, and reduced false positives on the rapid visual information processing task. Limitation: 45 of 107 enrolled did not complete, and the authors call for dose-ranging and longer trials.
• Chronic effects of Brahmi (Bacopa monnieri) on human memory — 76 adults aged 40 to 65, double-blind, randomized, placebo-controlled, three months of treatment, tested before, at three months, and six weeks after stopping. Measured retention of new information, learning rate, attention, verbal and visual short-term memory, retrieval of pre-existing knowledge, and questionnaire measures of everyday memory and anxiety. Bacopa improved a test of retention of new information; learning rate was unaffected, so the effect was a slower rate of forgetting. Attention, short-term memory, retrieval, everyday memory, and anxiety did not change. Limitation: the abstract does not state the dose or extract.
• Effects of 12-week Bacopa monnieri consumption on attention, cognitive processing, working memory, and functions of both cholinergic and monoaminergic systems in healthy elderly volunteers — 60 healthy older adults in Thailand (mean age 62.6; 23 men, 37 women), randomized double-blind placebo-controlled, 12 weeks of 300 mg or 600 mg of a standardized extract once daily or placebo, assessed every four weeks and four weeks after stopping. Measured working memory accuracy and reaction time, N100 and P300 event-related potential latencies and amplitudes, and plasma acetylcholinesterase and monoamine oxidase activity. Bacopa improved working memory, shortened N100 and P300 latencies, and suppressed plasma acetylcholinesterase activity. Limitation: the abstract does not separate the 300 and 600 mg results, and plasma enzyme activity is a proxy for what happens in the brain.
• Efficacy of standardized extract of Bacopa monnieri (Bacognize) on cognitive functions of medical students: a six-week, randomized placebo-controlled trial — 60 second-year medical students in Nagpur, India, randomized double-blind placebo-controlled parallel design, six weeks of 150 mg Bacognize extract twice daily or placebo, with a 15-day follow-up. Measured a battery of memory and cognitive tests plus blood biochemistry before and after. Statistically significant improvement on the cognitive tests with bacopa; serum calcium rose significantly but stayed within the normal range. Limitation: six weeks is the shortest positive trial, the abstract names neither the specific tests nor effect sizes, and the population is young adults studying full time.
• An acute, double-blind, placebo-controlled crossover study of 320 mg and 640 mg doses of a special extract of Bacopa monnieri (CDRI 08) on sustained cognitive performance — 24 healthy volunteers, double-blind placebo-controlled crossover, single doses of placebo, 320 mg, or 640 mg of CDRI 08 (KeenMind) on separate days. Measured six repetitions of the Cognitive Demand Battery after dosing plus cardiovascular measures and mood. The 320 mg dose improved performance at the first, second, and fourth repetitions post-dose; neither dose changed cardiovascular activity or blunted task-induced stress and fatigue ratings. Limitation: 24 people, a demanding and memory-light battery, and the 640 mg dose did not outperform 320 mg.
• An acute, double-blind, placebo-controlled cross-over study of 320 mg and 640 mg doses of Bacopa monnieri (CDRI 08) on multitasking stress reactivity and mood — 17 healthy volunteers, double-blind placebo-controlled crossover with 7-day washouts, single doses of placebo, 320 mg, or 640 mg of CDRI 08 (KeenMind), tested on a multitasking framework at baseline, 1 hour, and 2 hours. Measured cognitive outcomes from the multitasking battery, mood, and salivary cortisol. Change-from-baseline scores showed gains on the Letter Search and Stroop tasks at 1 and 2 hours after bacopa, some positive mood effects, and lower cortisol. Limitation: 17 people, and the authors state the effects need replication in a larger sample and outside a stressful test setting.
• The acute effects of an extract of Bacopa monniera (Brahmi) on cognitive function in healthy normal subjects — 38 healthy adults, double-blind placebo-controlled independent groups, a single 300 mg dose of bacopa extract (n=18) or placebo (n=20), tested before and 2 hours after. Measured a neuropsychological battery. No significant change on any test. Limitation: one dose, one extract, one time point, and the authors conclude only that there is no acute effect at the dose given.
• Safety evaluation of BacoMind in healthy volunteers: a phase I study — 23 healthy adult volunteers in a randomized, open-label, dose-escalation design, 30 days, one capsule of BacoMind daily at 300 mg for 15 days then 450 mg for 15 days. Measured clinical examination, hematology, blood biochemistry, and electrocardiograms before and after. No untoward changes in any measured parameter; mild gastrointestinal adverse events occurred and subsided on their own. Limitation: 30 days, 23 people, no placebo arm, and funded by the extract's manufacturer.
• A randomized controlled trial investigating the effect of Pycnogenol and Bacopa CDRI08 herbal medicines on cognitive, cardiovascular, and biochemical functioning in cognitively healthy elderly people: the Australian Research Council Longevity Intervention (ARCLI) study protocol — Protocol, not results. A randomized, placebo-controlled, double-blind three-arm trial planned to enroll 465 older adults for 12 months on CDRI 08 at 300 mg/day, Pycnogenol at 150 mg/day, or placebo, with cognitive, mood, cardiovascular, biochemical, and genetic assessments at baseline, 3, 6, and 12 months. The primary aim is cognitive performance over a year. A 2019 addendum added neuroimaging and gut-microbiome sub-studies. Limitation: the cognitive results had not appeared in PubMed when this page was built, so the trial is cited only to show that the long-term question is being asked and has not yet been answered.
• Relative efficacy of three medicinal plant extracts in the alteration of thyroid hormone concentrations in male mice — Male mice given Bacopa monnieri leaf extract at 200 mg/kg, compared with Aegle marmelos and Aloe vera extracts and controls. Measured serum T3 and T4, hepatic lipid peroxidation, and antioxidant enzyme activity. Bacopa raised serum T4 by 41% without raising hepatic lipid peroxidation; the other two extracts lowered thyroid hormones. Limitation: mice, a dose far above any human serving on a per-kilogram basis, and no human study has followed up, which is why this page lists thyroid medication as the key interaction to ask about.
• Protection from phenytoin-induced cognitive deficit by Bacopa monniera, a reputed Indian nootropic plant — Mice given phenytoin at 25 mg/kg orally for 14 days, with bacopa extract at 40 mg/kg added during the second week. Measured passive-avoidance learning and memory, maximal electroshock seizure protection, and locomotor activity. Phenytoin impaired passive-avoidance performance; bacopa reversed the impairment in both acquisition and retention without reducing phenytoin's anticonvulsant effect, and the effects were independent of motor stimulation. Limitation: mice and a 7-day co-administration; it establishes that the combination has been studied in animals, not that it is safe or useful in people.
• Neuropharmacological review of the nootropic herb Bacopa monnieri — Narrative review integrating animal, in vitro, and human trial literature on bacopa's mechanisms. Summarizes five proposed actions: antioxidant neuroprotection through redox and enzyme induction, acetylcholinesterase inhibition and/or choline acetyltransferase activation, amyloid reduction, increased cerebral blood flow, and modulation of acetylcholine, serotonin, and dopamine. Concludes that bacopa shows low toxicity in model organisms and humans but that long-term human toxicity studies have not been done. Limitation: a review, so it supports the mechanism section's framing rather than any single efficacy claim.

Dosage & forms

• Synapsa (CDRI 08):
• KeenMind (CDRI 08, Australian trade name):
• BacoMind:
• Bacognize:
• Generic extract, "50% bacosides":
• Whole-herb powder (dried leaf, capsules or loose):

Safety & interactions

How much to take

The research dose is 300 to 320 mg a day of a standardized extract, and the most common product dose, 300 to 500 mg, overlaps it only when the extract is one of the studied ones. The Swinburne trials used KeenMind at 300 mg or 2 x 150 mg (Stough 2001, Stough 2008) and CDRI 08 at 320 mg (Downey 2013, Benson 2014); the Oregon and Australian older-adult trials used 300 mg of a standardized extract and of BacoMind respectively (Calabrese 2008, Morgan 2010); the medical-student trial used 150 mg of Bacognize twice a day (Kumar 2016). A 320 mg serving of Synapsa is the trial dose by name. A 500 mg capsule of a generic "50%" extract is a bigger number and an unknown dose, because the standardization method does not match.

Take it once a day with a meal. Every trial that reported side effects reported gut complaints, and food is the standard mitigation; no trial compared morning with evening, so pick the meal you never skip. There is no required partner nutrient and no fat-solubility issue to plan around. The only thing to deliberately keep it away from is a medication on the interaction table below, until your prescriber has weighed in.

Judge it at 12 weeks, not before. The one trial that tested at both 5 and 12 weeks found the effects maximal at 12 (Stough 2001), and the 6-week Bacognize trial is the shortest with a positive result. If nothing has moved on a memory task you care about after three months of daily use, more time is unlikely to fix it.

Above 320 mg the picture is thin. A 30-day phase I study escalated from 300 to 450 mg with no safety signal (Pravina 2007), one 12-week trial used 600 mg (Peth-Nui 2012), and the acute studies used 640 mg once. The 2026 network meta-analysis found the working memory effect concentrated at 600 mg and above, but that comes from indirect comparison across trials rather than a dose-ranging study, and no trial has run 600 mg for longer than 12 weeks. Nobody has tested anything above 640 mg. No regulatory body has set an upper limit, because bacopa is not a nutrient; the practical ceiling is the highest dose a trial has actually used, and the practical floor is your stomach.

Safety profile

The strongest safety data are short and clean. In the phase I study, 23 healthy adults took BacoMind at 300 mg for 15 days and 450 mg for 15 more; clinical exams, blood counts, blood chemistry, and electrocardiograms were unchanged, and the only adverse events were mild gut complaints that resolved on their own (Pravina 2007). In the 12-week Oregon trial of adults 65 and older, the bacopa group logged 9 adverse events and the placebo group 10, "primarily stomach upset" (Calabrese 2008). The medical-student trial found serum calcium rose on Bacognize, within the normal range (Kumar 2016), a lab change worth knowing about rather than worrying about.

By frequency:

• Common: digestive. Increased stool frequency, abdominal cramps, and nausea, reported more often on bacopa than on placebo in the 98-person BacoMind trial (Morgan 2010); stomach upset in the Oregon trial (Calabrese 2008); mild gut events in the phase I study (Pravina 2007). Taking it with food is the fix in practice; if cramps persist past the first two weeks, stop.
• Uncommon: nothing else turned up consistently across the trials on this page. Heart rate drifted down over time in the Oregon trial's bacopa group while blood pressure did not change (Calabrese 2008).
• Rare: no published human case report of liver injury attributed to bacopa turned up in a PubMed search on the day this page was built; every bacopa-and-liver paper in the results was an animal study in which the extract protected the liver from a toxin. That is reassuring, not proof, and it is one reason not to combine bacopa with other botanicals of unknown provenance.

Drug interactions

Who should ask first

• Pregnant or nursing: no trial has included pregnant or nursing women, and there is no safety data either way. Skip it.
• On thyroid medication, a cholinesterase inhibitor, an anticholinergic, or any seizure medication: see the table; ask before starting.
• Scheduled for surgery: because of the cholinergic and possible sedative effects, tell the anesthesia team and stop ahead of the procedure on their schedule.
• Anyone with a sensitive gut: the side effect is predictable; start with food and be ready to stop.
• Children: the trials on this page are in adults. One pediatric trial exists; it is not the basis for anything here, and a pediatrician is the right first call.

Signs of too much

No toxic dose has been established and no overdose appears in the trial literature. The signal that you have overshot is the familiar one: cramping, loose stools, nausea. Cut the dose or stop. Nothing on this page suggests pushing through it.

If you take a medication whose dose a clinician adjusts to a lab value, bacopa is a conversation first and a capsule second.

Buying guide

The dose: 300 to 320 mg of a named, standardized extract

Match the label to a trial. The trial doses are 300 mg (KeenMind, BacoMind, and the standardized extract in the Oregon trial), 320 mg (CDRI 08), and 150 mg twice daily (Bacognize). A capsule that carries 320 mg of Synapsa is the Swinburne dose by name. A capsule that carries 500 mg of an unnamed extract is not a higher dose of the same thing; it is a different thing, and nobody has tested it.

The standardization: ask which method

"50% bacosides" means almost nothing without the assay. A UV spectrophotometry number counts every saponin and some things that are not saponins; an HPLC number counts named marker compounds and runs much lower for the same powder. The trial extracts are standardized by their makers to defined marker profiles, and CDRI 08 is specified at 55% bacosides with HPLC confirmation of its marker ratios. If a brand cannot tell you whether its percentage is UV or HPLC, assume UV and discount accordingly. The analytical section of our Synapsa review is the long version.

Proof of testing

Bacopa is a wetland plant, and wetland plants pick up what is in the water. Ask any brand for a certificate of analysis by lot that shows identity, bacoside potency, and heavy metals, and expect the numbers rather than a badge. Here is ours: Brain Advantage is made in our own cGMP facility in Mt. Angel, Oregon; every batch is tested in our in-house lab for identity, potency, and purity; and certificates of analysis are published by lot on our lab testing page. Brain Advantage was not yet posted there on the day this page was built, and until it is, the honest instruction is to ask us for the lot's certificate the same way you would ask anyone else. That is what to demand from every brand, including us. For the general framework for reading a label and a CoA, start with how to judge supplement safety and quality testing; for how often labels miss, what we found testing 44 supplements bought on Amazon is the uncomfortable read; and how we source and test raw botanicals covers the step before the capsule.

Fillers and the capsule

A standardized extract is bitter and bulky, so a two-capsule serving is normal and not a red flag. What is a red flag is a long "other ingredients" list: magnesium stearate is harmless but unnecessary, titanium dioxide in the shell is a choice a careful brand does not make, and a sugar-coated tablet is a sign the formulator cared more about the swallow than the dose. Vegetable cellulose capsules, a rice-derived flow agent, and silica are the short honest list.

Price per day

Divide the bottle price by the servings, not the capsules. Brain Advantage is $61.99 for 30 two-capsule servings, $2.07 a day, and that serving also carries 250 mg of Cognizin citicoline, 500 mg of lion's mane extract, 60 mg of ginkgo extract, and 30 mg of L-theanine, so the bacopa is one of five actives in the price. We did not price single-ingredient competitors for this page; when you do, the fair comparison is the per-day cost of 300 to 320 mg of a named extract, and a bottle that cannot name the extract has nothing to compare. Prices and serving counts checked on the live product page 2026-10-14.

Red flags

• A "nootropic blend" or "memory complex" that lists bacopa inside a proprietary blend with no milligram amount.
• "Clinically proven" with no study named. The trials on this page are the ones a label can honestly point to, and they belong to three named extracts.
• A bacoside percentage with no assay method, or a percentage above 55% with no explanation of how.
• "Bacopa leaf powder" at 500 mg sold on the strength of extract trials.
• Any bacopa product that does not tell you to take it with food.
• A label that leads with focus or energy. Memory retention is the finding; focus is the weakest one.

Where Brain Advantage fits

Brain Advantage carries 320 mg of Synapsa, standardized to 55% bacosides, in each two-capsule serving, alongside 250 mg of Cognizin citicoline, 500 mg of lion's mane mushroom extract, 60 mg of ginkgo leaf extract, 30 mg of L-theanine, and a short list of supporting nutrients, in vegetable cellulose capsules; 30 servings for $61.99, about $2.07 a day. The bacopa dose is the one the Swinburne trials used, by name and by amount. It is formulated to support memory, learning, and mental stamina in healthy adults with daily use, and the way to use it is the way the trials did: two capsules with a meal, every day, judged at twelve weeks.* If you want bacopa alone, look for 300 to 320 mg of one of the three named extracts and the same CoA standard. For where bacopa sits among the other ingredients people stack for mood and memory, our overview of supplements for mood, mind, and memory is the lateral read.

The bottom line

If you are a healthy adult who wants to hold on to what you learn, 300 to 320 mg a day of CDRI 08 (Synapsa or KeenMind), BacoMind, or Bacognize, taken with a meal for twelve weeks, has moderate trial support and a short, predictable side-effect list. If you want something to feel today, this is the wrong herb. If you take thyroid, seizure, or cholinergic medication, it is a conversation with your prescriber first. If your memory is fine and you are shopping for focus, the evidence says save your money or read the citicoline research instead.

Frequently Asked Questions

What is bacopa and what are bacosides?

Bacopa monnieri is a creeping wetland herb, called Brahmi in Ayurveda, whose extracts have been tested in placebo-controlled trials for memory in healthy adults. Bacosides are its active molecules: triterpenoid saponins, large soap-like compounds with sugar chains attached, of which the plant makes dozens (bacoside A3, bacopaside I and II, bacopasaponin C, and others). "Bacoside A" on a label is a group of at least four, not one compound. Extracts are standardized to a bacoside percentage, and the percentage depends on the assay: UV spectrophotometry counts every saponin, HPLC counts named markers and runs much lower for the same powder.

What did the memory trials show, and over how long?

Better retention of newly learned information after 12 weeks, measured as delayed recall on a word-list test. Four 12-week placebo-controlled trials found it: healthy adults on 300 mg of KeenMind (Stough 2001), 76 adults aged 40 to 65 (Roodenrys 2002), 54 adults 65 and older on 300 mg (Calabrese 2008), and 98 adults over 55 on 300 mg of BacoMind (Morgan 2010). A 90-day trial on 2 x 150 mg KeenMind found better spatial working memory (Stough 2008). The pooled attention effect across nine trials was 10 to 18 milliseconds faster reaction time (Kongkeaw 2014). Every one of those is a 6- to 13-week study; nothing longer has published results.

Why does bacopa take weeks to work?

Because the proposed mechanism is structural, not stimulant. In rats, bacopa extract raises the activity of enzymes that strengthen and grow synapses, especially in the hippocampus (reviewed in McPhee 2016), and building connections between neurons takes weeks. The human timing matches: a single 300 mg dose did nothing measurable at two hours (Nathan 2001), while the one trial that tested at both 5 and 12 weeks found the effect larger at 12 (Stough 2001). The only human mechanism measurement is a drop in blood acetylcholinesterase after 12 weeks (Peth-Nui 2012). Judge it at three months.

Is Synapsa (CDRI 08) different from generic bacopa extract?

Yes, in two ways that matter. Synapsa is CDRI 08, the extract developed at India's Central Drug Research Institute and sold as KeenMind in Australia; it is the material in the Swinburne trials (Stough 2001, Stough 2008, Downey 2013, Benson 2014), so its dose and effect are the published ones. It is also standardized to 55% bacosides with HPLC confirmation of specific marker ratios, where a generic "50%" is usually a UV number that counts non-bacoside saponins. A generic extract may be fine; it just cannot point to a trial, and its percentage is not comparable. [Our full review](/research/synapsa-cdri-08-bacopa-monnieri-cognitive-enhancement) covers the chemistry.

What dose did the trials use?

300 to 320 mg a day of a standardized extract, almost without exception. KeenMind at 300 mg or 2 x 150 mg (Stough 2001, 2008); CDRI 08 at 320 mg in the acute studies, with 640 mg also tested once (Downey 2013, Benson 2014); BacoMind at 300 mg (Morgan 2010) and up to 450 mg in a 30-day safety study (Pravina 2007); 300 mg of a standardized extract in Oregon (Calabrese 2008); 300 or 600 mg in Thailand (Peth-Nui 2012); and Bacognize at 150 mg twice daily (Kumar 2016). Nothing above 640 mg has been tested in people, and nothing above 600 mg for more than a single dose.

What are the digestive side effects and how do you avoid them?

More frequent stools, abdominal cramps, and nausea, reported more often on bacopa than on placebo in the 98-person BacoMind trial (Morgan 2010); "primarily stomach upset" in the Oregon trial, where adverse events were 9 on bacopa versus 10 on placebo (Calabrese 2008); and mild gut complaints that resolved on their own in the phase I study (Pravina 2007). Take it with a meal, once a day, at the trial dose rather than above it. If cramping persists past the first two weeks, stop; nothing in the trials rewards pushing through it.

Does bacopa interact with thyroid or other medications?

Possibly, and the human data are thin enough that the answer is "ask." In mice, 200 mg/kg of bacopa extract raised thyroxine (T4) by 41% (Kar 2002); no human trial has measured thyroid hormones on bacopa, so anyone on thyroid medication should tell the prescriber, especially before thyroid labs are drawn. Bacopa suppressed plasma acetylcholinesterase in a 12-week human trial (Peth-Nui 2012), which could add to cholinergic drugs and oppose anticholinergic ones. Additive drowsiness with sedatives is theoretical. In mice it reversed phenytoin's memory effects without blunting the drug (Vohora 2000), a research lead, not a reason to self-combine with seizure medication.

Can bacopa be taken with citicoline?

There is no trial of the pair, and no reason in the mechanisms to expect a conflict. Bacopa's proposed actions are slow and structural (synapse growth, acetylcholinesterase suppression); citicoline supplies choline and cytidine, the raw material for acetylcholine and for the phospholipids in neuron membranes, and its trials measured different outcomes on a different timescale. Stacking them is a bet on two mechanisms, not a documented interaction. [Our review of the Cognizin citicoline research](/research/cognizin-citicoline-neuropharmacology) covers what citicoline has shown on its own. Expect the bacopa half to need twelve weeks regardless of what it is paired with.

Who should not take bacopa?

Anyone pregnant or nursing, because no trial has included them. Anyone on thyroid medication, a cholinesterase inhibitor, an anticholinergic drug, or any seizure medication, until the prescriber has weighed in (the reasons are in the interaction table). Anyone about to have surgery, until the anesthesia team says when to stop. Children, without a pediatrician; the trials on this page are in adults. And, honestly, anyone whose memory is fine and who is shopping for focus or energy: the trials found memory retention, and the focus data are the weakest on the page. A healthy adult with no reason to expect a change is paying for a test score they will not notice.

What should a bacopa label show?

Five things. The extract's name (CDRI 08, Synapsa, KeenMind, BacoMind, or Bacognize) or, failing that, the maker. The milligrams per serving, outside any proprietary blend, at 300 to 320 mg. The bacoside percentage with the assay method, and a willingness to say HPLC. A certificate of analysis by lot showing identity, potency, and heavy metals, published rather than promised. And an instruction to take it with food. A label that leads with "focus" or "energy," hides the dose in a blend, or prints "clinically proven" without naming a study is telling you what it does not have.

Related Mt. Angel Vitamins Products

• Recommended for Bacopa Monnieri (Brahmi)

  Brain Advantage Nootropic For Memory, Focus And Mental Performance

  $61.99

  Shop Now