The short-acting hypnotic, triazolam 0.5 mg, was compared in a 7-day double-blind study with the longer acting drug, nitrazepam 5 mg. Of the 60 patients enrolled, 29 in each drug treatment group completed the study and were evaluated. All significant differences found favoured triazolam. On the first night, triazolam was found to be significantly more effective than nitrazepam in inducing and maintaining sleep and in increasing its overall duration. The patients' subjective preference, reflecting the depth of sleep experienced and the effect of the medication, reached highly significant levels in favour of triazolam. Later in the study the differences between the drugs disappeared although the depth of sleep as judged in the study was better with triazolam.