Reinstatement of responding maintained by cocaine or thiamylal.

Rhesus monkeys were trained to self-administer one of two reference drugs, either cocaine or thiamylal. The lowest dose of cocaine or thiamylal which maintained responding was determined. Responding extinguished when saline was substituted for a reference drug. Then the effects of a variety of other drugs on saline self-administration were determined. In the cocaine-trained monkeys, the i.m. administration of d-methylamphetamine, cocaine, morphine or codeine during a session reinstated responding for saline infusions; in contrast, naloxone, chlorpromazine, dimethyltryptamine, diazepam and secobarbital did not. Naloxone blocked responding produced by the i.m. administration of morphine. In the thiamylal-trained monkeys, the administration of secobarbital (i.m.), pentobarbital (p.o.), butabarbital (p.o.) and phenobarbital (p.o.) reinstated responding for saline. The onset of behavioral responding was related to the pharmacokinetic properties of the drugs, with phenobarbital greater than pentobarbital = butabarbital. In contrast, cocaine and d-amphetamine i.m. failed to reinstate the responding in monkeys which had been trained to self-administer thiamylal. These data support the hypothesis that drugs producing responding at a rate significantly greater than that produced by vehicle controls share reinforcing properties with the reference drug.