The effects of coffee and napping on nighttime highway driving: a randomized trial

Ann Intern Med. 2006 Jun 6;144(11):785-91. doi: 10.7326/0003-4819-144-11-200606060-00004.


Background: Sleep-related accidents often involve healthy young persons who are driving at night. Coffee and napping restore alertness, but no study has compared their effects on real nighttime driving performances.

Objective: To test the effects of 125 mL of coffee (half a cup) containing 200 mg of caffeine, placebo (decaffeinated coffee containing 15 mg of caffeine), or a 30-minute nap (at 1:00 a.m.) in a car on nighttime driving performance.

Design: Double-blind, randomized, crossover study.

Setting: Sleep laboratory and open highway.

Participants: 12 young men (mean age, 21.3 years [SD, 1.8]).

Measurements: Self-rated fatigue and sleepiness, inappropriate line crossings from video recordings during highway driving, and polysomnographic recordings during the nap and subsequent sleep.

Intervention: Participants drove 200 km (125 miles) between 6:00 p.m. and 7:30 p.m. (daytime reference condition) or between 2:00 a.m. and 3:30 a.m. (coffee, decaffeinated coffee, or nap condition). After intervention, participants returned to the laboratory to sleep.

Results: Nighttime driving performance was similar to daytime performance (0 to 1 line crossing) for 75% of participants after coffee (0 or 1 line crossing), for 66% after the nap (P = 0.66 vs. coffee), and for only 13% after placebo (P = 0.041 vs. nap; P = 0.014 vs. coffee). The incidence rate ratios for having a line crossing after placebo were 3.7 (95% CI, 1.2 to 11.0; P = 0.001) compared with coffee and 2.9 (CI, 1.7 to 5.1; P = 0.021) compared with nap. A statistically significant interindividual variability was observed in response to sleep deprivation and countermeasures. Sleep latencies and efficiency during sleep after nighttime driving were similar in the 3 conditions.

Limitations: Only 1 dose of coffee and 1 nap duration were tested. Effects may differ in other patient or age groups.

Conclusions: Drinking coffee or napping at night statistically significantly reduces driving impairment without altering subsequent sleep.

Publication types

  • Randomized Controlled Trial

MeSH terms

  • Adult
  • Automobile Driving*
  • Caffeine
  • Circadian Rhythm / physiology
  • Coffee*
  • Cross-Over Studies
  • Double-Blind Method
  • Fatigue / etiology
  • Fatigue / physiopathology
  • Fatigue / prevention & control*
  • Humans
  • Male
  • Polysomnography
  • Self Concept
  • Sleep Deprivation / complications
  • Sleep*
  • Task Performance and Analysis


  • Coffee
  • Caffeine