Background: Cognitive impairments are now recognized in Parkinson’s disease. Some of these deficits owe to disease pathology itself whereas others are due to paradoxical effects of dopaminergic medications, such as levodopa. The dopamine overdose hypothesis proposes that dissimilar effects of medication on cognition depend on baseline endogenous dopamine levels in underlying brain regions. We sought to directly test this prevalent theory. Methods: We tested healthy adults, who presumably have optimal endogenous dopamine levels, in two sessions. Participants received 100/25 mg of levodopa/carbidopa in one session and an equal volume of placebo in the other. During each session, participants completed a probabilistic reversal learning task. The number of trials to task completion was used as a behavioural proxy of learning performance. Results: A paired t-test covaried with drug-placebo order revealed that healthy adults learned more poorly on levodopa compared to placebo. Conclusions: Our findings suggest that baseline endogenous dopamine levels are a critical factor determining the effects of dopaminergic medications on cognition, independent of Parkinson’s disease pathology. Partitioning which cognitive functions are helped versus hindered by medication and improving our understanding of the underlying psychopharmacology of these effects is important for improving treatment strategies in Parkinson’s disease.