The effects of the anorectics benzphetamine, chlorphentermine, clortermine, mazindol, phendimetrazine, and phenmetrazine on food intake were compared to the effects of d-amphetamine in rhesus monkeys given daily access to food pellets. The ability of these compounds to maintain intravenous self-administration under a fixed-ratio 10 schedule was also determined in rhesus monkeys. All drugs reduced food intake in a dose-related manner. d-Amphetamine was the most potent. Mazindol, chlorphentermine, phenmetrazine, and phendimetrazine were approximately 1/5 to 1/9 as potent as d-amphetamine while benzphetamine and clortermine were 1/14 and 1/20 as potent, respectively. Benzphetamine, mazindol, and phenmetrazine were self-administered above saline levels and were approximately equipotent. Chlorphentermine and clortermine were not self-administered and phendimetrazine was self-administered by only one of four monkeys at one dose. Thus, although all of the compounds were effective anorectics, chlorphentermine, clortermine, and phendimetrazine did not function as positive reinforcers. Since the ability of a drug to function as a positive reinforcer is related to its dependence potential, these three compounds might be less subject to abuse when used therapeutically. Within the group of 3 compounds that was self-administered, benzphetamine was relatively more potent as a positive reinforcer than as an anorectic. Therefore, this drug might be a less desirable compound for therapeutic use.