Previous studies have shown that bright light (2500 lux) suppresses nocturnal secretion of melatonin, while dim light (500 lux) has little or no effect. We have studied the effect of varying intensities of light on 5 normal male volunteers (age 18-28). The experiment was divided into 3 parts which took place at weekly intervals. Subjects remained under artificial light (fluorescent strip 150-250 lux) between 2000 h-2300 h, they then retired to bed in darkness. On each occasion, between 0030 h and 0100 h, the subjects were required to get up and were treated with light of different intensities; (a) less than 1 lux, (b) 300 lux and (c) 2500 lux respectively. Subjects returned to bed in darkness until 0700 h. Blood was sampled hourly from 2000 h-1000 h with additional samples at 2330 h, 0015 h, 0030 h, 0045 h, 0115 h and 0130 h. Plasma melatonin and 6-sulphatoxymelatonin (aMT6s), the major melatonin metabolite, were measured by radioimmunoassay. Dim (300 lux) and bright (2500 lux) light, both significantly suppressed melatonin levels compared to less than 1 lux (P less than 0.05 and P less than 0.01 respectively) at the following time points 0100 h, 0115 h and 0130 h. One subject did not show suppression with 300 lux. There was also a significant suppression of aMT6s levels, compared to less than 1 lux, after both 300 lux and 2500 lux at 0115 h (P less than 0.05, P less than 0.01), 0130 h (P less than 0.01, P less than 0.01) and 0200 h (P less than 0.01, P less than 0.001) respectively.(ABSTRACT TRUNCATED AT 250 WORDS)