Melatonin output covaries with sleepiness, with both peaking early morning. Bright white light suppresses this output, but it is not known if such treatment ameliorates nighttime sleepiness during sleep deprivation. However, sleep-deprived subjects find such light irritating. Humans are particularly sensitive to green light, and melatonin output is more readily suppressed by this hue. A pilot study using different green light regimens showed that sleep-deprived subjects well tolerated 2,000 lux green light given 10 min hourly, and that this dose reduced nighttime melatonin output. The main study gave this light treatment vs. a low intensity red/green light control, from 1900 hr for 11 hr, to two groups of subjects (n = 6 each) sleep deprived for 36 hr. Urine was collected at 6-hr intervals during sleep loss and on a baseline day. Vigilance performance, subjective sleepiness, and oral temperature were monitored during sleep loss. The experimental condition suppressed urine 6-hydroxymelatonin sulfate (aMT6s) output between 0000 hr and 0600 hr, and increased it 0600-1200 hr; but there was no change in total 24-hr values. The control condition had no such effects. The oral temperature rhythm remained unchanged. Vigilance and subjective sleepiness were improved significantly relative to control values during 0000-0600 hr; these improvements were maintained somewhat over the 0600-1200-hr period, contrary to what one might expect if the delayed melatonin surge at this time was increasing sleepiness. Although the bright green light helped counteract sleepiness, any causal link with changes in melatonin output seem tenuous.