The effects of triazolam (0.25 mg) or flurazepam (15 mg) were evaluated in 13 elderly (ages 64-79) insomniacs. Subjects were reasonably healthy, ambulatory, and complained of disturbed sleep. Sleep apnea syndromes were ruled out by nocturnal polysomnogram. Sleep, daytime sleepiness [Multiple Sleep Latency Test (MSLT) and Stanford sleepiness scale (SSS)], performance, and mood [Profile of Mood States (POMS)] were assessed on five consecutive days. Placebo was given on nights 1 and 2; active medications were given on nights 3-5. Sleep time was increased by approximately 1 h in both groups. MSLT showed increased sleepiness with flurazepam and decreased sleepiness with triazolam. MSLT scores were unrelated to nocturnal sleep parameters in the flurazepam group and showed a pattern of correlation, though nonsignificant, in the triazolam group. Vigilance was impaired with flurazepam and unchanged with triazolam. Other performance tests showed slight improvement or no change with drugs. Mood tended to be improved after flurazepam ingestion and unchanged after triazolam. These findings suggest that, although both compounds improve nocturnal sleep in elderly insomniacs, there is significant residual sedation with flurazepam and improved daytime alertness with triazolam. Neither compound had a significant effect on nocturnal respiration in these non-sleep-apneic elderly subjects.