Rats were trained and tested in a step-down inhibitory avoidance task (0.3-mA footshock). Training-test interval was 6 h. In Experiment 1, animals received, 1 h before training, an ip injection of vehicle or diazepam (2.0 mg/kg) and, 30 s after training and/or 30 min prior to testing, ip saline, epinephrine (6.25 micrograms/kg or 125.0 micrograms/kg), naloxone (0.5 mg/kg), or beta-endorphin (1 micrograms/kg). In the vehicle-pretreated animals, post-training epinephrine (6.25 micrograms/kg) and naloxone enhanced, and post-training beta-endorphin and epinephrine (125.0 micrograms/kg) reduced, retention test performance; and pretest beta-endorphin and epinephrine (125.0 micrograms/kg) reversed the latter effect and enhanced retention on their own. Diazepam lowered memory scores on its own and prevented all other drug effects with the exception of post-training facilitation by epinephrine (6.25 micrograms/kg). In previous papers it was shown that post-training facilitation by epinephrine is due to an influence on storage processes, whereas all the other drug effects described above result from the post-training establishment of state dependency to either beta-endorphin or epinephrine, and therefore to a process involving further acquisition and storage. The present findings suggest that diazepam selectively hindered the acquisition and/or storage processes involved in state dependency. This conclusion is strengthened by the findings from Experiment 2, which showed, using a classic 2 x 2 design, that diazepam itself did not induce state dependency but, rather, depressed acquisition and/or storage of the avoidance task.