Under sleep-laboratory control, the efficacy of flurazepam hydrochloride (15 mg) was evaluated in 6 women (age range, 67-82 years) with objectively verified insomnia. A 15-night, single-blind, crossover procedure was followed. Sleep records obtained during 3 placebo-baseline nights, 7 consecutive flurazepam nights, and 3 placebo-withdrawal nights were evaluated by means of electroencephalographic, electro-oculographic, and electromyographic criteria. A statistically significant reduction in sleep latency and total awake time and a corresponding increase in total sleep time (P less than 0.05) were demonstrated during the active drug period. No evidence of diminishing effectiveness was observed during the 7 days of drug administration. For the rapid-eye-movement (REM) stage, a significant decrease (P less than 0.05) in mean REM percent was noted during the drug period despite an increase in mean absolute REM time. No REM rebound occurred upon drug withdrawal. There were no significant changes in mean percentages for stages 3 and 4 during the drug period and the withdrawal period. Adverse reactions were rare (chiefly some daytime drowsiness in 2 subjects).