Rationale <p>The probabilistic reversal learning task (PRL) is sometimes used in the context of major depressive disorder (MDD) to assess impairments in cognitive flexibility and feedback sensitivity because behavior in the task is sensitive to pharmacological interventions. Because traditional antidepressants are limited in their effectiveness, new drugs are needed to combat MDD. Ketamine has recently been investigated in the context of probabilistic reversal learning (PRL), but findings regarding its therapeutic efficacy have been mixed. One reason for this could be that almost all non-human versions of the PRL use signaled reward omission (i.e., timeout) as the punishing stimulus. It has long been known that timeout periods do not always function as punishers, and the inclusion of a known effective punishing stimulus could help to produce results of improved translational value.</p> Objective <p>The present experiment sought to examine the effects of ketamine in the PRL when electric footshocks accompanied timeout periods or not.</p> Methods <p>A baseline of PRL performance was established with 40 rats in which typical timeouts followed non-rewarded trials. In Phase 2, half the rats also received probabilistic footshock punishment for non-rewarded trials, while the other half continued under baseline conditions. Finally, a single dose of ketamine was administered to half of the rats (<i>n</i> = 10) in each condition (i.e., shock and no shock).</p> Results <p>Shock punishment increased behavioral persistence and cognitive flexibility in the PRL, but ketamine had no effect beyond causing acute impairments.</p> Conclusion <p>These results suggest that the conditions of punishment during the PRL can have a significant impact on performance in the task and corroborate previous findings that ketamine may not impact cognitive flexibility or reward processing in healthy rats.</p>