The bottom line: how much and when you drink caffeine matters most
The clearest answer from the research is that afternoon caffeine can harm your sleep, but the damage is dose- and timing-dependent. A 2023 meta-analysis of 24 studies—the largest and most comprehensive analysis here—found that caffeine consumption reduces total sleep time by an average of 45 minutes and sleep efficiency by 7% [1]. It also increases the time it takes to fall asleep (sleep onset latency) by 9 minutes and the amount of time you spend awake after initially falling asleep (wake after sleep onset) by 12 minutes [1]. To put that in perspective, a 45-minute loss of sleep is like missing almost an entire sleep cycle, which can leave you groggy the next day.
The same study provides practical guidance: to avoid reductions in total sleep time, a standard cup of coffee (about 107 mg of caffeine) should be consumed at least 8.8 hours before bedtime, and a stronger pre-workout supplement (217.5 mg) at least 13.2 hours before [1]. This means that a 3 PM coffee could still be affecting your sleep if you go to bed at 11 PM, since that's only 8 hours later—right at the edge of the safe window. A 4 PM coffee would be too late.
The dose-response relationship: more caffeine, closer to bed, equals worse sleep
A 2023 study that directly tested different doses and timings in 23 healthy men confirms this pattern. When participants consumed 400 mg of caffeine (roughly four cups of coffee) four hours before bedtime, their total sleep time dropped by 54 minutes, sleep efficiency fell by 10.2%, and they spent 25.3 more minutes awake after falling asleep [3]. In contrast, a lower dose of 100 mg (about one cup) consumed 12 hours before bed had a much smaller effect [3]. This shows a clear dose-response: the higher the dose and the closer to bedtime, the greater the disruption. The study also found that caffeine reduced deep sleep (N3) by 4.55% and increased light sleep (N1 and N2) by 5.5%, meaning you get less of the restorative sleep your body needs [3].
Another 2024 study on sleep-deprived individuals found that even when people were exhausted, caffeine (2.5 mg/kg, about 175 mg for a 70 kg person) taken 6.5 hours before recovery sleep still reduced total sleep time by 30 minutes and deep sleep (N3) by 35.6 minutes [2]. This suggests that caffeine's sleep-disrupting effects persist even when you're very tired, which is relevant for shift workers or anyone catching up on sleep after a late night.
The caveat: habitual drinkers may be less affected
Not everyone responds the same way. A 2021 study of 20 young men who were regular, moderate caffeine consumers (up to 300 mg/day) found that when they took 150 mg of caffeine three times daily (morning, midday, and afternoon) for 10 days, their nighttime sleep structure—total sleep time, sleep latency, and sleep architecture—did not differ significantly from when they took a placebo [4]. This suggests that if you're a habitual coffee drinker, your body may adapt, and afternoon caffeine may not strongly impair your sleep quality or duration.
However, this study did find a subtle effect: caffeine reduced EEG power density in the sigma frequency range (12-16 Hz) during non-REM sleep, which may indicate early signs of overnight withdrawal from the stimulant [4]. So while your sleep might look normal on the surface, there could be underlying changes in brain activity. Also, this study used a specific schedule (last caffeine dose 8 hours before bed) and healthy young men, so the results may not apply to everyone, especially older adults or those with sleep disorders.
The real-world data from a 2023 study of financial traders adds another layer: caffeine consumption led to a reduction in sleep quantity (not quality), while alcohol affected sleep quality [5]. Interestingly, when both were used together, alcohol partially masked caffeine's effect on sleep quantity, suggesting that people may self-medicate with alcohol to counteract caffeine's stimulation—a risky cycle [5].
About These Sources
This answer is built on 5 peer-reviewed studies — published from 2021 to 2024, 1 from 2024 or later, 3 in Q1 journals, collectively cited 309 times — selected as the most relevant from 5 studies that passed quality screening, drawn from 54 papers retrieved from a database of over 500 million.
Sources used in this answer
The effect of caffeine on subsequent sleep: A systematic review and meta-analysis
In a meta-analysis of 24 studies, caffeine reduced total sleep time by 45 minutes and sleep efficiency by 7%, and increased sleep onset latency by 9 minutes and wake after sleep onset by 12 minutes; to avoid sleep loss, coffee (107 mg) should be consumed at least 8.8 hours before bed, and a pre-workout supplement (217.5 mg) at least 13.2 hours before [1].
Caffeine Intake Alters Recovery Sleep after Sleep Deprivation
In a double-blind crossover trial with 41 participants, acute caffeine (2.5 mg/kg) taken 6.5 hours before recovery sleep after 38 hours of sleep deprivation reduced total sleep time by 30 minutes and deep sleep (N3) by 35.6 minutes, and increased nighttime awakenings [2].
O050 The Dose and Timing Relationship between Caffeine and Subsequent Sleep
In a randomized study of 23 healthy men, 400 mg of caffeine consumed 4 hours before bedtime reduced total sleep time by 54 minutes and sleep efficiency by 10.2%, while 100 mg consumed 12 hours before had minimal effects, showing a clear dose-response relationship [3].
The impact of daily caffeine intake on nighttime sleep in young adult men
In a double-blind crossover study of 20 habitual caffeine consumers, 450 mg of caffeine daily (in three doses) consumed in the morning and afternoon did not significantly impair total sleep time, sleep latency, or sleep architecture compared to placebo, though it reduced EEG sigma power during non-REM sleep [4].
Sleep, alcohol, and caffeine in financial traders
In a real-world study of financial traders over six weeks, caffeine consumption reduced sleep quantity (not quality), while alcohol reduced sleep quality; when combined, alcohol partially mitigated caffeine's effect on sleep quantity, suggesting a self-medication cycle [5].