The present investigation assessed whether increased congruency between ACTH state present shortly after training and that at testing contributed to memory recovery. If recovery were related to an increased correspondence between internal state present after training and that at testing, then suppressing ACTH release should block memory recovery. This was the hypothesis that was examined in the present investigation. Specifically, animals were trained on a passive avoidance task, administered hypothermia (the amnestic agent) and, shortly prior to testing, given treatments known to be effective in reversing memory loss induced by hypothermia. Before training (Experiment 1) or testing (Experiment 2) animals were injected with either dexamethasone (an agent that suppresses ACTH release) or saline. Results, in general, indicated that when ACTH release was suppressed, a blunted recovery effect was obtained. This reduction in the extent of memory recovery was observed when ACTH was suppressed either at training or at testing. These data are interpreted as providing support for an ACTH-related, state-dependent retention mechanism contributing to recovery from hypothermia-induced retrograde amnesia in rats.