We studied the sleep effect, the carry over effect (sleepiness) on polysomnography (PSG), the sleep propensity test (SPT), the subjective sleep rating scale for the ultra short acting hypnotics triazolam and short acting hypnotics brotizolam. Ten healthy male volunteers were used for double blind crossover design by randomized allocation with a single oral administration. Informed consent was obtained from all subjects. Placebo (PL), triazolam (TL, 0.125 mg), triazolam (TH, 0.25 mg) and brotizolam (BR, 0.25 mg) were administered at 11 p.m. PSG was recorded from 11 p.m. to 7 a.m. in the next morning. Then, Stanford Sleepiness Scale (SSS) and Kwansei Gakuin Sleepiness Scale (KSS) were checked before 8 a.m., and SPT was recorded at 8:00–8:20 a.m. In PSG of total sleep, TH decreased the frequency of stage shifts significantly ( p<0.05) compared to PL. TH and BR increased the percentage of stage 1 combined with stage 2 sleep significantly ( p<0.01) and decreased the percentage of REM sleep significantly ( p<0.01) compared to PL. However, all the drugs had no effect on the percentage of slow wave sleep compared to PL. Although all drugs had no effect on the subjective feeling of sleepiness in SSS and KSS, BR decreased sleep latency significantly ( p<0.05) compared to PL in SPT.