Coffee and Memory: Does Caffeine Help You Remember?
Author: Ukrainian Psychological Hub · Published: September 28, 2026 · Editorial Policy
Coffee can help memory under some conditions, but it is not a reliable general memory enhancer. The strongest and most consistent cognitive effect of caffeine is improved alertness and attention, not a universal improvement in remembering. A major review of caffeine and memory concluded that results vary substantially with the kind of memory being tested, dose, timing, habitual use, sleep, age, and other experimental factors. The 2021 review by Zhang and Madan found no single, uniform caffeine effect across memory systems.
That distinction matters because feeling more awake can make it easier to pay attention to material, and better attention can improve what gets encoded in the first place. But a person can feel sharper without showing better recall or recognition later. Reviews of caffeine and performance consistently find more dependable benefits for vigilance, attention, reaction time, and alertness than for memory or higher-order executive functions. McLellan, Caldwell, and Lieberman’s review describes this pattern directly, and a 2025 systematic review and meta-analysis of attention trials found small acute improvements in both attention accuracy and reaction time. For the dedicated attention question, see Coffee and Focus: Does Caffeine Really Improve Concentration?. For the broader cognition framework spanning attention, reaction time, memory, and executive function, see Coffee and Cognition: Attention, Reaction Time, Memory, and Executive Function.
So the best evidence-based answer to “Does coffee help you remember?” is: sometimes, indirectly and task-dependently; occasionally directly in specific experimental conditions; and sometimes not at all. A famous 2014 experiment found better 24-hour mnemonic discrimination after 200 mg of caffeine taken after learning, but a 2020 follow-up found no detectable enhancement and ruled out medium or large benefits, while a 2025 randomized study found worse accuracy in a face-recognition consolidation task. The evidence is therefore much more interesting than the slogan “caffeine improves memory.”
Coffee and memory: the quick answer
If coffee helps you stay alert enough to read, listen, practice, or retrieve information effectively, it can support learning indirectly. This is especially plausible when baseline alertness is low. In a coffee-based experiment with college-age adults, caffeinated coffee improved explicit memory in the early morning but not in the afternoon, suggesting that the effect depended on time of day and physiological state rather than representing a permanent increase in memory capacity. Sherman and colleagues’ 2016 study is one reason the question cannot be answered with a simple yes or no.
For long-term memory consolidation, the evidence is mixed. Borota and colleagues reported in Nature Neuroscience that caffeine given immediately after learning improved performance 24 hours later on a task requiring discrimination between previously seen images and similar lures. But Aust and Stahl’s 2020 follow-up found no caffeine effect and concluded that, if an enhancement exists, it is substantially smaller than the original result suggested. A 2025 randomized double-blind study using face recognition found lower correct-choice rates and more false alarms after 200 mg of post-encoding caffeine.
For working memory, results are also inconsistent. A small 2024 randomized double-blind trial in habitual coffee and tea drinkers found that 200 mg of caffeine impaired spatial working-memory performance while leaving verbal working memory unchanged. That trial is available through PubMed. Small studies should not be treated as universal rules, but this result is a useful counterexample to the idea that caffeine improves every form of memory.
Coffee is not the same thing as caffeine
Coffee is a beverage containing hundreds of compounds; caffeine is one psychoactive component of that beverage. Most mechanistic claims about acute alertness and many cognition experiments concern caffeine specifically. Some studies administer pure caffeine, while others compare caffeinated and decaffeinated coffee. Those designs answer related but not identical questions.
This matters in everyday life because the amount of caffeine in coffee varies widely with serving size, bean species, coffee dose, recipe, extraction, and beverage type. A “cup of coffee” is not a standardized pharmacological dose. The FDA lists a broad range of about 113–247 mg for a 12-fluid-ounce serving of regular brewed non-specialty coffee and emphasizes product variation. Current FDA caffeine guidance provides those reference values. Our complete guide to caffeine in coffee explains why brew method, serving size, and coffee composition change total dose.
That is also why research using a 200 mg caffeine capsule cannot be translated into “two cups of coffee will improve memory.” The experimental dose is precise; a real-world coffee serving often is not. The source, timing, habitual use, and task also differ.
What researchers mean by memory
Memory is not one mental container. Research separates multiple processes and systems, and caffeine can affect them differently. Encoding is the initial processing of new information. Consolidation refers to processes that stabilize a memory after learning. Retrieval is the act of accessing stored information later. A result at one stage cannot automatically be generalized to the others.
Researchers also distinguish working memory from longer-term episodic or declarative memory. Working memory is the limited-capacity system used to hold and manipulate information over seconds while performing a task. Episodic memory concerns events and experiences. Recognition asks whether something has been encountered before. Recall requires generating information without the same level of external cueing. Implicit memory can influence behavior without deliberate recollection.
These distinctions explain why apparently conflicting caffeine studies can all be valid. Caffeine may speed responses on one task, improve explicit recall in a low-alertness condition, show no benefit on delayed free recall, and impair a particular spatial or face-discrimination measure. Those findings describe different cognitive operations.
How caffeine could affect remembering
Adenosine blockade changes arousal and attention
Caffeine’s best-established acute mechanism is antagonism of adenosine receptors. Adenosine signaling contributes to sleep pressure; blocking its receptors promotes wakefulness and changes neural activity related to arousal. McLellan and colleagues’ review summarizes adenosine receptor blockade as the central mechanism behind caffeine’s familiar alerting effects.
The memory implication is often indirect. If you are drowsy, lapsing in attention, or responding slowly, caffeine can make it easier to remain engaged with the material. Better engagement can improve encoding because information that is not adequately attended to is less likely to be learned well. This is an attention pathway into memory, not proof that caffeine strengthens every memory trace.
Feeling more alert is not the same as remembering more
Subjective alertness and objective memory can diverge. You can feel energized, motivated, or mentally fast while delayed recall remains unchanged. This is one reason a study session can feel more productive after coffee even when a later memory test shows little difference.
Expectations can also shape the experience. In a study that crossed actual caffeine with information about whether participants were supposedly receiving caffeine, caffeine shortened reaction time in a memory-scanning task but did not improve delayed free recall, and the expectancy manipulation did not produce a memory benefit. Oei and Hartley’s experiment illustrates why subjective confidence, arousal, and memory performance should be measured separately.
Consolidation is a separate question
A study that gives caffeine before learning cannot easily separate a direct memory effect from better attention during encoding. That methodological problem motivated the 2014 post-study experiment: participants learned first and received caffeine afterward. The design was intended to isolate consolidation more cleanly. The original Borota study reported an inverted-U-shaped dose-response pattern and interpreted the 24-hour advantage as a consolidation effect rather than a retrieval effect.
That study became influential because it suggested something more specific than ordinary stimulation. Yet replication and extension work has complicated the picture. Aust and Stahl detected no enhancement under their design, and Leon and colleagues found impaired face-recognition accuracy after post-encoding caffeine. The responsible conclusion is that post-learning caffeine is not an established general-purpose consolidation enhancer.
What the evidence shows by memory type
Attention and encoding: the most reliable pathway
The most reproducible cognitive benefit of caffeine is attention. The 2025 meta-analysis by Kløve and Petersen included 31 randomized double-blind placebo-controlled trials with 1,455 healthy rested adults and found small improvements in both reaction time and accuracy. Higher doses produced larger reaction-time effects, while accuracy followed a nonlinear pattern and declined beyond a point. That last finding is important: more stimulation is not equivalent to better cognitive performance.
An earlier review of double-blind placebo-controlled attention studies similarly concluded that caffeine has clear beneficial effects on attention. Einöther and Giesbrecht’s review also argued that the attention effects could not be explained entirely by reversal of withdrawal. This matters for learning because attention is part of successful encoding, though an attention benefit still does not guarantee better delayed memory.
Working memory: no universal boost
Working memory is often advertised as if it were a single score that caffeine reliably raises. The literature does not support that simplification. Different tasks place different demands on verbal storage, spatial maintenance, updating, inhibition, and response speed. Caffeine can change some components without improving the overall ability to hold and manipulate information.
The 2024 study of habitual coffee and tea drinkers is a concrete example: spatial working memory worsened after 200 mg of caffeine, while verbal working memory did not significantly change. Read the randomized trial. Its sample was only 18 participants, so it should be treated as a specific experimental result rather than a population rule.
Explicit memory: benefits appear state-dependent
Some evidence suggests caffeine can help explicit memory when baseline arousal is suboptimal. In the 2016 coffee study, young adults who drank caffeinated rather than decaffeinated coffee performed better on explicit memory in the early morning, but caffeine did not improve explicit memory in the afternoon and did not improve implicit memory. The authors interpreted the effect as specific to participants’ non-optimal time of day.
This is a useful model for real life. Coffee may be most helpful when it moves someone from sleepy and disengaged toward an effective level of alertness. That is different from saying that caffeine continuously raises memory above an already well-rested, attentive baseline.
Long-term retention and mnemonic discrimination: mixed evidence
The best-known positive result is the 2014 post-study consolidation experiment. Participants receiving 200 mg after encoding showed better next-day discrimination between targets and similar lures than placebo, while 100 mg did not show the same benefit and a larger dose did not produce a larger improvement. The published study describes an inverted-U response.
But the later follow-up is essential to the story. Aust and Stahl’s 2020 study did not replicate the caffeine enhancement and statistically ruled out medium and large improvements under its conditions. The authors also raised the possibility that differences in the original literature could partly reflect caffeine withdrawal in control conditions.
The 2025 face-recognition study makes the evidence even more task-specific. Participants received 200 mg immediately after viewing faces and were tested the next day. Instead of improved consolidation, the caffeine group showed lower correct-choice rates and more false alarms in discriminating studied faces from similar new ones. The Scientific Reports article therefore argues against treating 200 mg as a universal memory-consolidation dose.
Naps, caffeine, and memory are not interchangeable
A direct comparison of caffeine, a daytime nap, and placebo found a striking dissociation. In the experiment, a 60–90 minute nap improved recall on verbal memory tasks relative to caffeine and placebo; caffeine did not reproduce the same pattern and also impaired some non-declarative learning measures. Mednick and colleagues’ 2008 study is useful because it demonstrates that wake-promoting stimulation and memory-supporting sleep are different interventions.
Does caffeine help you study?
Coffee can be useful for studying when the main problem is sleepiness, fading vigilance, or difficulty sustaining attention. That is the domain where caffeine has its most consistent evidence. It may help you stay engaged long enough to read, solve problems, rehearse material, or perform retrieval practice effectively.
It is much less defensible to use caffeine as though it were a chemical substitute for learning. Memory depends on what you do with the material: organizing it, retrieving it, spacing practice over time, connecting it with prior knowledge, and sleeping after learning. Coffee can alter the state in which those activities occur; it cannot make ineffective study strategies effective by itself.
The practical implication is to evaluate coffee by the bottleneck it addresses. If the bottleneck is low alertness, caffeine may help. If the bottleneck is poor comprehension, passive rereading, interference, insufficient practice, or sleep loss, another cup does not directly solve the problem.
Why late-night study coffee can become a memory trade-off
Using caffeine to extend studying into the night can create a conflict between immediate alertness and later memory. Caffeine can make a tired person feel more capable of continuing, but sleep itself contributes to learning and memory. If the same caffeine dose delays sleep, shortens sleep, or reduces sleep quality, the next-day benefit may be smaller than the extra study time feels in the moment.
A meta-analysis of total sleep deprivation found detrimental effects on memory when a night of sleep deprivation occurred either before learning or after learning, with the larger pooled effect when deprivation happened before encoding. Newbury and colleagues’ meta-analytic review supports a simple principle: sleep is part of the memory process, not unused time after learning.
Caffeine itself can disrupt subsequent sleep. A systematic review and meta-analysis of 24 studies concluded that caffeine reduces total sleep time and sleep efficiency and delays sleep onset, with the timing needed to avoid measurable sleep effects depending on dose. Weakley and colleagues’ 2023 meta-analysis provides the sleep evidence. Our dedicated Coffee and Sleep guide covers dose, timing, deep sleep, and individual sensitivity in detail.
The European Food Safety Authority notes that even a single 100 mg dose may affect sleep duration and patterns in some adults, particularly when consumed close to bedtime. EFSA’s caffeine safety page is a useful official reference. For memory, the point is not that every late coffee will impair learning; it is that the full equation includes both the hours you stay awake and the sleep you preserve afterward.
How much caffeine improves memory?
There is no established universal “memory dose” of caffeine. The number 200 mg appears frequently because several experiments used it, including the positive 2014 consolidation study, the 2020 follow-up that did not replicate a meaningful enhancement, the 2024 spatial working-memory study that found impairment, and the 2025 face-recognition study that found worse accuracy. The same dose produced different outcomes because the tasks, participants, timing, and study designs differed.
That is why the 200 mg result should not be converted into a personal dosing instruction. An experimental dose identifies what happened in that experiment; it does not establish an optimal dose for an individual reader. The broader cognition literature shows that dose-response curves can differ across outcomes. The recent attention meta-analysis, for example, found continued reaction-time gains at higher doses while accuracy improved only to a point and then declined. See the 2025 attention meta-analysis.
For general safety rather than memory optimization, the FDA cites 400 mg per day as an amount not generally associated with negative effects for most adults, while emphasizing individual sensitivity and total intake from all sources. FDA guidance also lists insomnia, anxiety, jitters, palpitations, nausea, and headache among possible signs of excessive caffeine. EFSA likewise states that intakes up to 400 mg per day generally do not raise safety concerns for healthy adults other than pregnant women and gives a lower 200 mg per day level for pregnancy. EFSA guidance. Safety limits are not cognitive-performance targets.
Habitual coffee use, tolerance, and withdrawal
Habitual use complicates caffeine research because a regular user who arrives at a laboratory after several caffeine-free hours may already be experiencing withdrawal. If caffeine restores that person toward their normal state, an apparent “enhancement” can partly reflect withdrawal reversal rather than performance above their usual caffeinated baseline.
This remains a genuine methodological debate. A review by Jack James argued that much of the apparent performance and mood benefit in habitual users can be explained by withdrawal reversal. James’s 2005 review represents that position. In contrast, Einöther and Giesbrecht’s attention review concluded that caffeine’s attention effects could not be attributed entirely to withdrawal reversal. The 2021 caffeine-and-memory review also identified withdrawal, tolerance, diet, and sleep controls as recurring methodological problems. See Zhang and Madan.
For an individual coffee drinker, this means the morning experience may combine several things at once: pharmacological stimulation, relief of withdrawal symptoms, a learned routine, taste and aroma cues, expectation, and the ordinary rise in alertness after waking. Those components can all feel like “my memory works better after coffee” even when the measurable effect on memory is narrower.
Expectation, confidence, and the psychology of feeling cognitively sharper
Coffee is unusually rich in cues: smell, warmth, bitterness, the cup itself, the place where it is consumed, and a long history of pairing those cues with work or study. These cues can create expectations before caffeine reaches peak physiological effect. Expectations can influence subjective arousal and some performance measures, but they should not be equated with memory enhancement.
That sensory-cue pathway is different from caffeine’s effect on cognitive performance. For the dedicated autobiographical-memory explanation, see Coffee and Nostalgia: Why Smell and Taste Can Trigger Powerful Memories.
In the Oei and Hartley experiment, actual caffeine shortened reaction time in memory scanning, but neither the message about receiving caffeine nor caffeine itself improved delayed free recall. The study separates expectancy from delayed memory performance. This distinction is psychologically important: confidence that one is “switched on” can be real as an experience without proving that more information will be remembered tomorrow.
The same principle applies to coffee rituals. A familiar study coffee can mark the transition into a work state and reduce friction around beginning a task. That behavioral effect may improve how much studying gets done. It remains different from a direct biochemical increase in memory storage.
Can too much coffee make memory worse?
Caffeine can plausibly hurt memory performance when its other effects interfere with the task. Excessive arousal, shakiness, racing thoughts, gastrointestinal discomfort, or anxiety can compete with attention. The FDA explicitly lists anxiety, jitters, insomnia, palpitations, nausea, and headache among signs that caffeine intake may be too high for a person. See the FDA’s current guidance.
The cognitive evidence also includes direct negative results. The 2024 habitual-user trial found worse spatial working memory after 200 mg, and the 2025 face-recognition study found more false alarms and lower accuracy after the same nominal dose. Those studies do not prove that ordinary coffee generally harms memory; they show why “more caffeine equals more memory” is scientifically untenable.
Anxiety is another reason the subjective experience matters. A person who becomes tense or overstimulated after coffee may find it harder to concentrate on the relevant material even if caffeine is increasing physiological arousal. Our Coffee and Anxiety guide separates caffeine dose, panic sensitivity, ordinary jitteriness, and individual differences.
Individual differences: why the same coffee can produce different cognitive experiences
People differ in habitual caffeine exposure, sleep debt, time of day, baseline arousal, body size, medication use, anxiety sensitivity, pregnancy status, metabolism, and the size and concentration of the beverage. These factors change both caffeine exposure and the context in which memory is tested.
The 2021 memory review emphasized demographic and methodological moderators, including age, sex, dose, tolerance, metabolism, diet, and sleep. Zhang and Madan’s review therefore argues against treating a single laboratory result as a universal cognitive law.
Coffee preference adds another layer. Someone may drink coffee because they enjoy the flavor, because it is part of a morning identity, because it structures a study ritual, or because stopping it causes withdrawal symptoms. None of these motives tells us whether that person’s episodic or working memory improves after caffeine. The experience and the measured cognitive effect must be kept separate.
Coffee, aging, dementia, and long-term memory
The question “Does coffee help memory?” often expands into a different question: “Does long-term coffee drinking protect against cognitive decline or dementia?” These are not the same research problem. Acute experiments test what happens over minutes or hours after caffeine. Dementia studies usually follow dietary habits and health outcomes over years or decades.
A 2020 systematic review of 61 studies described the epidemiologic evidence as inconclusive overall, with positive associations depending on dose, caffeine source, sex, and other study characteristics. Chen and colleagues’ systematic review suggested that moderate caffeine consumption may be associated with lower risk in some groups but did not establish caffeine as a dementia-prevention treatment.
More recently, a large 2026 prospective cohort study in JAMA followed 131,821 participants from the Nurses’ Health Study and Health Professionals Follow-up Study for up to 43 years. Higher caffeinated-coffee intake was associated with lower dementia risk and somewhat more favorable cognitive outcomes, while decaffeinated coffee was not significantly associated with dementia risk. The JAMA study is important because of its size and long follow-up, but it remains observational: association does not prove that coffee caused the lower risk.
These long-term findings should therefore not be used to promise that drinking coffee will prevent Alzheimer’s disease, reverse cognitive decline, or preserve an individual’s memory. They add to a developing observational literature that deserves continued study.
Common myths about coffee and memory
Myth: coffee is a memory drug
The evidence is much stronger for alertness and attention than for memory. Caffeine may support memory indirectly by improving the state in which learning occurs, and certain experiments report memory benefits, but effects are not consistent across memory systems or conditions.
Myth: 200 mg after studying will lock in what you learned
That claim overgeneralizes one famous result. The original 2014 study reported a benefit on mnemonic discrimination, but the 2020 follow-up found no meaningful enhancement and the 2025 face-recognition trial found worse discrimination accuracy. Post-learning caffeine is therefore an open, task-dependent research question rather than a validated study technique.
Myth: if coffee improves focus, it must improve recall
Attention and memory are connected but separable. Better attention can improve encoding, yet later recall can remain unchanged. Reaction time can also improve without better accuracy or retention.
Myth: stronger-tasting coffee has a stronger memory effect
Taste intensity does not reveal total caffeine dose. Dark, bitter, or concentrated-tasting coffee can feel “strong” for sensory reasons, while total caffeine depends on bean composition, dry coffee dose, extraction, beverage volume, and serving size. Our caffeine-in-coffee guide separates concentration, total dose, and subjective effect.
Myth: more caffeine produces better memory
There is no linear memory rule of this kind. Cognitive outcomes can show plateaus, null effects, or impairment. The attention meta-analysis even found a nonlinear accuracy pattern at higher doses. Increasing dose also increases the chance of sleep disruption and other unwanted effects.
Myth: coffee prevents dementia
Long-term cohort studies can identify associations, and the 2026 JAMA study reported a favorable association for caffeinated coffee. That evidence does not establish coffee as a preventive treatment, and earlier systematic reviews describe substantial heterogeneity.
Practical implications for learning and remembering
Use coffee for the problem it is most likely to solve: low alertness. If you are adequately rested and already attentive, the expectation of a large memory boost is not well supported. If you are sleepy, caffeine may help you engage with learning more effectively, but it still matters what learning strategy you use.
Keep the distinction between caffeine amount and coffee size in view. A small espresso and a large brewed coffee can have very different caffeine concentrations and total doses. When dose matters, estimate caffeine rather than using the beverage name as a proxy. The CF70 caffeine guide provides the broader coffee-dose framework.
Protect sleep after learning. A study session that is extended by caffeine but followed by delayed or shortened sleep can create a trade-off. If remembering tomorrow is the goal, the sleep consequences of late caffeine belong in the decision. See Coffee and Sleep for the evidence on timing and sensitivity.
Do not treat a research dose as a personal prescription. The memory literature does not establish an optimal individual dose. If caffeine produces anxiety, palpitations, tremor, gastrointestinal symptoms, or sleep problems, those effects are relevant even when a laboratory study used a similar dose without difficulty.
For exams and demanding learning, the most defensible role for coffee is supportive rather than magical: it can help maintain attention, especially when alertness is low, while effective learning methods and adequate sleep do the deeper work of building and stabilizing memory.
What the evidence does and does not allow us to conclude
Established evidence: caffeine acutely increases alertness and, on average, improves several measures of attention and reaction time in healthy adults. It can also disrupt sleep, depending on dose, timing, and individual sensitivity.
Supported but conditional evidence: caffeine can improve performance on some memory tasks under particular circumstances, including some low-alertness or time-of-day conditions. A post-learning consolidation benefit has been reported in a prominent trial, but it has not emerged as a robust general effect across follow-up studies and tasks.
Evidence against a universal benefit: controlled studies include null and negative findings for working memory, delayed memory, mnemonic discrimination, and face recognition. The effect cannot be summarized as “caffeine improves memory.”
Observational evidence: higher long-term caffeinated-coffee intake has been associated with lower dementia risk and some favorable cognitive outcomes in large cohorts, including a major 2026 JAMA study. These associations do not establish causality or justify using coffee as a dementia treatment.
Frequently asked questions
Does coffee improve short-term memory?
Sometimes on particular tasks, but not consistently. Caffeine more reliably improves alertness and attention. A 2024 trial, for example, found impaired spatial working memory and no significant change in verbal working memory after 200 mg in habitual coffee and tea drinkers.
Does caffeine improve long-term memory?
The evidence is mixed. A 2014 randomized study found improved 24-hour mnemonic discrimination after post-study caffeine, but a 2020 follow-up did not replicate a meaningful benefit, and a 2025 face-recognition study found worse discrimination accuracy.
Should I drink coffee before studying?
Coffee may help if low alertness or sleepiness is limiting your ability to concentrate. It is not established as a general memory enhancer, and the benefit can be offset if caffeine is taken late enough to disrupt sleep.
Should I drink caffeine after studying to improve consolidation?
There is not enough consistent evidence to recommend post-study caffeine as a general memory-consolidation strategy. The key positive study has been followed by null and negative findings in related paradigms.
Is 200 mg of caffeine best for memory?
No universal best memory dose has been established. The same nominal 200 mg dose has produced positive, null, and negative memory findings in different experiments.
Does decaf coffee improve memory?
Decaf contains much less caffeine, so it does not provide the same pharmacological exposure as regular coffee. Coffee cues and expectations can still influence subjective experience, but that is different from demonstrating a memory benefit.
Can coffee help me remember during an exam?
It may help attention or alertness, particularly if you are below your normal alertness level. Whether that translates into better recall depends on what was learned, the memory task, prior sleep, habitual caffeine use, and individual response.
Can coffee make me forgetful?
Ordinary coffee is not established as a cause of general memory loss. But excessive caffeine can interfere with sleep or produce anxiety and overstimulation, and some controlled studies report worse performance on specific memory tasks.
Does coffee prevent Alzheimer’s disease?
No causal preventive effect has been established. Large observational studies have reported associations between caffeinated coffee and lower dementia risk, but observational associations cannot prove that coffee prevents Alzheimer’s disease or other dementias.
Is caffeine better than a nap for memory?
Not necessarily. In one direct comparison, a 60–90 minute nap improved verbal recall relative to caffeine and placebo. Caffeine is effective for wakefulness; sleep and napping can support memory through different mechanisms.
Why does coffee make me feel smarter even when memory does not improve?
Caffeine can increase alertness, reduce sleepiness, speed reaction time, and change subjective energy. Those effects are noticeable immediately. Delayed memory is a different outcome, so the feeling of mental sharpness can be stronger than the measurable memory change.
Does stronger coffee improve memory more?
No. Sensory strength is not a measure of caffeine dose, and more caffeine does not produce a reliable linear memory benefit. Total dose, timing, sleep, tolerance, and the memory task all matter.
