The discriminative stimulus (DS) and subjective effects of two anorectic drugs, phenylpropanolamine (PPA) and mazindol (MAZ), were studied in a group of normal, healthy adults trained to discriminate between placebo and 10 mg d-amphetamine (AMP). Of 20 subjects who underwent discrimination training, 12 (discriminators) reliably learned the AMP-placebo discrimination. Each discriminator was tested with two doses of PPA (25 and 75 mg) and two doses of MAZ (0.5 and 2.0 mg) to determine whether the DS effects of these drugs would substitute for those of AMP. The high dose of each drug produced primarily (approximately 80%) drug-appropriate responding, whereas the low dose of each drug resulted in primarily placebo-appropriate responding. The subjective effects of PPA were a biphasic function of dose, with 25 mg producing mild sedative-like effects and 75 mg producing stimulant-like effects similar to, but weaker than, those obtained with AMP. MAZ, on the other hand, produced only a few changes in mood (increased anxiety, decreased hunger). Thus, although both PPA and MAZ substituted for AMP in terms of discrimination responding, only PPA produced AMP-like subjective effects. These results provide evidence for a dissociation between the subjective effects (as measured by self-report questionnaires) and the DS effects of drugs in humans.