The conventional wisdom about caffeine and sleep is simple: avoid it in the afternoon if you want to sleep well at night. But new research on individual variation in caffeine metabolism is revealing that the right cutoff time is highly personal.

Caffeine’s half-life—the time it takes for your body to eliminate half of the caffeine you consumed—averages around five to six hours. That’s the basis for the common recommendation to stop consuming caffeine by early afternoon. If you have coffee at 2pm, there’s still roughly a quarter of that caffeine in your system at bedtime.

But “averages around five to six hours” hides enormous individual variation. Some people metabolize caffeine in three hours. Others take nine hours or longer. This difference is largely genetic, determined primarily by variations in the CYP1A2 gene that produces the enzyme responsible for breaking down caffeine.

If you’re a fast metabolizer, having coffee at 4pm might have negligible impact on your sleep. If you’re a slow metabolizer, even morning coffee could still be affecting you at bedtime, and afternoon caffeine is almost certainly disrupting your sleep.

The catch is that most people don’t know which category they fall into. You can get genetic testing that includes CYP1A2 variants, but it’s not commonly done. So we’re left making general recommendations that fit average people moderately well but miss the mark for people at either end of the distribution.

Recent sleep studies that tracked caffeine consumption and sleep quality using wearables and sleep diaries found something interesting: self-reported caffeine sensitivity correlates reasonably well with actual sleep impact. People who say they’re sensitive to caffeine generally are, and people who claim it doesn’t affect them often show no measurable sleep disruption.

This suggests that personal experience is a decent guide. If you’ve been drinking coffee at 3pm for years and sleeping fine, you’re probably a fast metabolizer. If afternoon caffeine leaves you tossing and turning, you’re probably a slow metabolizer. Trust your own observations over generic advice.

But there’s a complication. Caffeine tolerance develops with regular use. People who consume caffeine daily become less sensitive to its stimulant effects. This tolerance is real, but it’s primarily tolerance to the subjective feeling of being wired, not tolerance to caffeine’s actual physiological effects on sleep architecture.

Studies measuring sleep in habitual caffeine users show that even when people feel like caffeine doesn’t affect them, it still reduces slow-wave deep sleep and can delay sleep onset. You might fall asleep normally and feel like you slept fine, but objective measures show your sleep quality was reduced.

This creates a perceptual problem. If you don’t feel like caffeine is affecting your sleep, you might continue consuming it late in the day, not realizing it’s degrading sleep quality in ways you’re not consciously aware of.

Age matters too. Older adults tend to metabolize caffeine more slowly on average, and they’re also more sensitive to sleep disruption. The afternoon coffee that didn’t bother you in your twenties might be a problem in your fifties.

Interactions with medications can alter caffeine metabolism. Some common drugs—including certain antibiotics, antidepressants, and birth control pills—slow caffeine breakdown. If you start a new medication and suddenly find yourself more sensitive to caffeine, this interaction might be why.

The type of caffeine source matters less than you might think. Coffee, tea, energy drinks, and caffeine pills all deliver the same molecule. But they differ in caffeine content per serving and in the presence of other compounds that might modulate effects.

Tea contains L-theanine, which may partially counteract caffeine’s jittery effects. Some people find they can drink tea later in the day than coffee without sleep impact. This could be L-theanine, or it could just be that tea typically has less caffeine per cup than coffee.

Decaf coffee isn’t caffeine-free. It typically contains 5-15mg per cup, compared to 80-100mg in regular coffee. For most people, this amount is negligible. For highly sensitive individuals or slow metabolizers, even decaf could have measurable effects if consumed close to bedtime.

The research on optimal caffeine timing is difficult because controlled experiments can’t perfectly simulate real-world conditions. Lab studies that give subjects caffeine at specific times and measure sleep in sleep labs find significant sleep disruption from afternoon and evening caffeine. But these studies often use relatively large doses and may not reflect how people actually consume caffeine.

Field studies that track real-world behavior are messy because people differ in so many ways—not just caffeine metabolism, but baseline sleep quality, stress levels, other substance use, and countless other factors that affect sleep.

What’s emerging from recent research is a more nuanced picture. For people with normal caffeine metabolism, the cutoff is probably around 6-7 hours before bedtime for moderate doses. For a 10pm bedtime, that means stopping around 3-4pm.

Fast metabolizers can likely extend that. Slow metabolizers should be more conservative—possibly stopping at noon or even limiting themselves to morning caffeine only.

Individual experimentation is the most practical approach. Try eliminating all afternoon caffeine for two weeks and see if you notice sleep improvements. Then carefully reintroduce it at progressively later times and observe the effects.

Wearable sleep trackers can help, though they’re not perfect. If your deep sleep percentage or sleep efficiency drops on days when you had afternoon caffeine, that’s useful information even if the absolute numbers aren’t laboratory-accurate.

Sleep debt complicates the picture. When you’re chronically under-slept, you might rely on afternoon caffeine to function, which then further disrupts sleep, creating a vicious cycle. Breaking that cycle often requires enduring a few days of fatigue while you eliminate late caffeine and allow sleep to normalize.

Shift workers and people with irregular schedules face additional challenges. The standard advice assumes a conventional sleep-wake schedule. If you’re sleeping during the day and working at night, caffeine timing needs to be adjusted accordingly based on when you sleep, not what the clock says.

Some sleep physicians are starting to recommend personalized caffeine guidelines based on detailed sleep history and caffeine response rather than one-size-fits-all rules. This makes sense but requires more conversation and patient experimentation than simply saying “no caffeine after 2pm.”

The practical takeaway is that generic caffeine timing advice is a reasonable starting point, but individual variation means you need to figure out what works for you specifically. If you’re struggling with sleep and consuming caffeine past early afternoon, trying an earlier cutoff is a low-cost experiment worth attempting.

For most people, that experiment will reveal that they’re more caffeine-sensitive than they realized. Sleep improvements from earlier caffeine cutoffs are common once people actually try it consistently. But for genuine fast metabolizers, strict early cutoffs might be unnecessarily restrictive.

Pay attention to your sleep, be honest about caffeine’s effects, and be willing to adjust based on results rather than assumptions. The right answer isn’t the same for everyone, and that’s okay.