The disposition of chlordiazepoxide, diazepam, oxazepam and lorazepam was studied before and after disulfiram administration to normal subjects and alcoholic patients. Similar studies to determine the effect of ethanol administration on diazepam and oxazepam disposition were also performed. After disulfiram, decreases in the plasma clearance of chlordiazepoxide (54%, p < 0.05; diazepam, 41% p < 0.05) and their active metabolites were observed. Oxazepam and lorazepam have no important active metabolites and the effect of disulfiram on their net disposition was minimal. In normal subjects ethanol (> 800 mg/l maintained for 8 h) increased the AUC free diazepam (26.6 +/- 17.3% mean +/- SD, p < 0.05) and decreased the AUC for N-desmethyldiazepam (50.5 +/- 11.7%, p < 0.05) indicating inhibition of diazepam N-desmethylation. In contrast, there was no significant change in the disposition of oxazepam in four healthy subjects after a single oral dose of ethanol. Thus, disulfiram and ethanol appear not to inhibit the metabolic disposition of oxazepam or lorazepam, benzodiazepine derivatives principally biotransformed by glucuronidation. On pharmacokinetic grounds oxazepam and lorazepam may be the drugs of choice if benzodiazepine therapy is required in chronic alcoholics.