An emerging phenomenon in our understanding of the pathophysiology of mental illness is the idea that specific proteins may form insoluble aggregates in the brains of patients, in partial analogy to similar proteinopathies in neurodegenerative diseases. Several proteins have now been detected as forming such aggregates in the brains of patients, including DISC1, dysbindin-1 and TRIOBP-1. Recently, neuronal PAS domain protein 3 (NPAS3), a known genetic risk factor for schizophrenia, was implicated through a V304I point mutation in a family with major mental illness. Investigation of the mutation revealed that it may lead to aggregation of NPAS3. Here we investigated NPAS3 aggregation in insular cortex samples from 40 individuals, by purifying the insoluble fraction of these samples and testing them by Western blotting. Strikingly, full-length NPAS3 was found in the insoluble fraction of 70% of these samples, implying that aggregation is far more widely spread than can be accounted for by this rare mutation. We investigated the possible mechanism of aggregation further in neuroblastoma cells, finding that oxidative stress plays a larger role than the V304I mutation. Finally, we tested to see if NPAS3 aggregation could also be seen in blood serum, as a more accessible tissue than the human brain for future diagnosis. While no indication of NPAS3 aggregation was seen in the serum, soluble NPAS3 was detected, and was more prevalent in patients with schizophrenia than in those with major depressive disorder or controls. Aggregation of NPAS3 therefore appears to be a widespread and multifactorial phenomenon. Further research is now needed to determine whether it is specifically enhanced in schizophrenia or other mental illnesses.