Parietal cortex is causally required for state-dependent decisions.

During perceptual decision-making, behavioral performance varies with internal states like arousal, motivation, and strategy, but how these fluctuations impact distributed neural processes remains unclear. We trained mice on a navigation-based sound-localization task and examined how behavioral performance state changes relate to neural processing in cortex. Hidden Markov modeling shows that expert mice alternate between a high-performance state with optimal stimulus weighting and low-performance states characterized by high choice bias. Optogenetic inactivation of the posterior parietal cortex (PPC) decreases task accuracy and increases occupation of biased performance states. Two-photon calcium imaging shows that performance states strongly influence population activity in the PPC but not the primary auditory cortex. Individual PPC neurons are better explained by external inputs during biased performance, while shared neural variability is strongest during high-performance states. These findings demonstrate that parietal cortex activity is causally required for decisions and is linked to behavioral performance states.