In a residential hospital ward setting, either sodium pentobarbital, diazepam or ethanol was made available for oral ingestion to volunteer human subjects with documented histories of drug abuse. During specified portions of the day, tokens could be earned by riding an exercise bicycle and exchanged for doses of a drug. Increases in the required minimum interingestion interval from 0-30 min produced decreases in the number of ingestions of sodium pentobarbital and diazepam. In another experiment, increases in the dose per ingestion (30-90 mg of sodium pentobarbital, 2-10 mg of diazepam or 1.86-11.14 g of ethanol) produced increases in the number of ingestions. In both experiments, the effects of the manipulated variable were similar for all the drugs studied. Human self-administration research with the sedative drugs, sodium pentobarbital and diazepam is feasible, for which substantive experimental data have not previously been reported. Both dose and minimum interingestion interval seem to bear a systematic controlling relationship to the occurrence of drug self-administration behavior.