Rats were tested using a lever pressing/feeding procedure in which a preferred food (Bioserve pellets) was available by pressing a lever on a fixed ratio 5 schedule, but a less preferred food (lab chow) was also available concurrently in the operant chamber. The effects of repeated (14 day) injections of haloperidol, clozapine and thioridazine were compared. Haloperidol (0.05-0.15 mg/kg) significantly reduced lever pressing and increased chow intake throughout the drug treatment period. Injections of clozapine (2.0-6.0 mg/kg) suppressed lever pressing but failed to produce substantial increases in chow intake. In the haloperidol experiment there was a significant inverse correlation between lever pressing and chow intake, but in the clozapine experiment there was not. Regression analysis indicated that rats treated with the high dose of clozapine showed some tolerance to the suppression of lever pressing. Tests of sedation also were conducted before and after the instrumental behavior sessions. Haloperidol produced little or no sedative effect in the dose range tested. Clozapine produced substantial sedation during the first 10 days of administration, but this effect, like the suppression of lever pressing, showed signs of tolerance. Thioridazine (3.0-9.0 mg/kg) produced some effects that resembled haloperidol, and other effects, including sedation, that resembled clozapine. These studies indicate that haloperidol suppresses lever pressing for food at low doses that do not produce severe motivational or sedative effects that disrupt food intake. In contrast, it appears as though the suppression of lever pressing produced by clozapine stems from a sedative effect that also serves to set limits on chow intake. These results indicate that haloperidol and clozapine suppress lever pressing through different mechanisms.