- A discrete-trial choice procedure was used to examine the reinforcing and subjective effects of four anorectic drugs (mazindol, benzphetamine, phenylpropanolamine and phenmetrazine) in groups of normal healthy adults. For each experiment, subjects first sampled placebo and a dose of one of the drugs (mazindol: 0.5, 1.0 and 2.0 mg; benzphetamine: 25 and 50 mg; phenylpropanolamine: 12.5, 25 and 50 mg; phenmetrazine: 25 and 50 mg; all p.o.). Subjects were then allowed to choose between drug and placebo on five separate occasions. The relative frequency with which active drug was chosen over placebo was used as an index of the drug's reinforcing efficacy. Subjective effects were measured with an experimental version of the Profile of Mood States, a short form of the Addiction Research Center Inventory and a series of visual analog scales. The rank order for reinforcing efficacy was benzphetamine approximately phenmetrazine greater than placebo greater than phenylpropanolamine much greater than mazindol. Ratings of drug liking were positively correlated with number of drug choices for each drug. Benzphetamine and phenmetrazine produced subjective effects characteristic of amphetamine-like drugs and increased ratings of drug liking. Mazindol produced only dysphoric subjective effects and decreased ratings of drug liking. Phenylpropanolamine had no significant effects on subjective measures or drug-liking ratings. These findings are consistent with the presumed dependence potential of these compounds, and demonstrate the validity of this experimental paradigm for assessing the reinforcing effects of anorectics in normal human volunteers.