Inconsistency among beliefs is a hallmark of irrationality. Despite longstanding interest in inconsistency in philosophy and psychology, empirical evidence of synchronically held inconsistencies among people’s belief has proven elusive. Here, across two pre-registered experiments (Ns = 500, 274), we identify inconsistent beliefs simultaneously held by individual participants. Drawing on Sommer et al.’s (2023) proposal that accessibility in memory helps people achieve consistent beliefs, we constructed sets of questions that facilitated or hindered the accessibility of relevant knowledge. Our results support the proposal that consistency is enforced when beliefs are simultaneously accessible, rather than resulting from exhaustive consistency-checking. We find that when participants have simultaneous access to inconsistent beliefs–even regarding inconsequential general knowledge topics–they tend to revise their beliefs toward consistency. Furthermore, we experimentally distinguish an alternative explanation that the inconsistencies we evoked are merely inconsistent responses. Taken together, our results suggest that inconsistency among beliefs may be common, arising when inconsistencies are inaccessible.