Sixty insomniac patients participated in a controlled double-blind parallel group study designed to investigate the dose-response relationship of zopiclone. Following 1 day of treatment with placebo, patients were randomly assigned to 1 of 6 groups and received treatment for 7 days with either placebo, or flurazepam 30 mg, or zopiclone, 3.75 mg, 7.5 mg, 11.25 mg or 15 mg. Four patients were dropped from the study; two from the placebo group due to ineffectiveness and one each in zopiclone 11.25 mg and 15 mg groups due to side-effects. Flurazepam 30 mg significantly improved sleep induction and maintenance by comparison to placebo and was indistinguishable from zopiclone 7.5 mg or higher. Results of a self-administered sleep questionnaire found a predominantly linear relationship between the dose of zopiclone administered and the degree of sleep improvement. The greatest increment in improvement was generally obtained with 3.5 mg and 7.5 mg of zopiclone, with some additional benefit occurring with zopiclone 11.25 mg. Clinicians' global impressions showed that the severity of illness clearly decreased in a dose related manner up to zopiclone 11.25 mg. Although zopiclone was well tolerated at 3.75 mg and 7.5 mg, an increase in side-effects occurred at 11.25 mg and 15 mg, which favours the use of 7.5 mg zopiclone as the optimum dose for most patients, although certain patients may benefit from a higher dose of the drug when well tolerated.