Within a residential research ward, five human volunteers with histories of sedative drug abuse were exposed to progressive-ratio schedules of pentobarbital (200, 400, 600 mg) or placebo self-administration. All doses were letter-coded and administered under double-blind conditions. To obtain a single letter-coded dose, three subjects were required to press a set of buttons a specified number of times and two subjects were required to ride a stationary bicycle for a specified period of time. Only one dose could be obtained per day and the button-pressing or riding requirement for each letter-coded dose was increased over successive sessions until subjects failed to meet the progressive-ratio requirement (i.e., the subject chose not to work for the dose). Drug-effect ratings and subjective measures were taken 2 h after drug administration. Pentobarbital maintained dose-related increases in the maximum progressive-ratio requirement completed, the subject and staff ratings of drug effect, the subject ratings of drug 'liking', and the scores on the PCAG scale of the Addiction Research Center Inventory. The present study suggests that progressive-ratio schedules are sensitive and valid procedures for providing information about the relative reinforcing efficacy of drugs.