A double-blind crossover study was performed to evaluate the efficacy of hypnotics in geriatric insomniacs. Twenty-seven patients with a mean age of 70 years (range 60-94 years) received each of five treatments on 5 consecutive nights. The treatment conditions, consisting of chloral hydrate 250 and 500 mg, triazolam 0.25 and 0.50 mg, and placebo, were administered using a Latin Square design. Subjective estimates of sleep were collected in the morning following each treatment night. The patients' global evaluation of effectiveness indicated that triazolam 0.25 mg and 0.50 mg improved sleep more than placebo, while chloral hydrate 250 and 500 mg were not better than placebo. Triazolam 0.50 mg was felt to be significantly better than either dose of chloral hydrate. In addition, triazolam 0.50 mg was found to significantly decrease the patients' estimates of their sleep latency. Patients estimated their total sleep time to be longer following the use of triazolam 0.25 mg as compared to choral hydrate 500 mg, and their estimates of the number of awakenings was significantly lower on triazolam 0.50 mg than it was on chloral hydrate 500 mg or placebo.