Twenty-six healthy subjects from 19 to 85 yr old took single 15-mg doses of flurazepam (FLZ). Concentrations of desalkylfurazepam (DAFLZ), its principal metabolite, were measured by gas-liquid chromatography in multiple samples drawn 7 or more days after the dose. For the first 6 to 8 hr after drug, several additional FLZ metabolites appeared in plasma, but only DAFLZ was detected from 12 hr onward. Its elimination half-life (t1/2) (range, 37 to 289 hr) was longer in elderly than in young men (mean 74 and 160 hr, p less than 0.05), but t1/2 in young and elderly women was much the same (90 and 120 hr, P = NS). Eighteen of the 26 subjects then received FLZ, 15 mg, nightly for 15 consecutive nights. Blood samples were drawn during FLZ dosage and in the withdrawal period, and morning self-ratings of mood and sleep patterns were obtained using visual analogue scales. DAFLZ cumulation was extensive, with a mean cumulation ratio of 7.5. Mean steady-state plasma levels of DAFLZ were higher in elderly than in young men (81 and 53 ng/ml, P less than 0.05), but values were essentially the same in elderly and young women (85 and 86 ng/ml). Single-dose t1/2 correlated with washout t1/2 after termination of FLZ treatment (r = 0.87, P less than 0.01). Clinical self-ratings indicated increases over time in perception of morning sedation; changes slowly reverted to baseline in the week after dosage. Sleep patterns also improved on FLZ (shortened latency, longer duration, "deeper" sleep). After termination of treatment, sleep parameters returned to baseline with a suggestion of "overshoot" sleep disturbance at days 5 and 7 after drug. There was no evidence of increased sensitivity to FLZ in the elderly. Subjects did not perceive any impairment of intellectual function or motor performance, and no other adverse reactions were reported.