Aims: To evaluate the association of relapse and healthcare resource utilization in patients with schizophrenia (SZ), bipolar disorder (BD), or major depressive disorder (MDD) who switched antipsychotic medication versus those who did not.Materials and methods: Medicaid claims from six US states spanning six years were retrospectively analyzed for antipsychotic switching versus non-switching. For all patients with SZ, BD, or MDD, and for the subset of patients who also had ≥1 extrapyramidal symptoms (EPS) diagnosis at baseline, times to the following outcomes were analyzed: underlying disease relapse, other psychiatric relapse, all-cause emergency room (ER) visit, all-cause inpatient (IP) admission, and EPS diagnosis.Results: Switchers (N = 10,548) had a shorter time to disease relapse, other psychiatric relapse, IP admissions, ER visits, and EPS diagnosis (all, log-rank p < .001) than non-switchers (N = 31,644). Switchers reached the median for IP admission (21.50 months) vs non-switchers (not reached) and for ER visits (switchers, 9.07 months; non-switchers, 13.35 months). For disease relapse, other psychiatric relapse, and EPS diagnosis, <50% of patients had an event during the two-year study period. Subgroup analysis of those with ≥1 EPS diagnosis revealed similar associations.Limitations: Only association, not causation, may be inferred, and there may be differences between the patient groups in parameters not evaluated.Conclusions: These results show that disease and other psychiatric relapse, all-cause ER visits, IP admissions, and EPS diagnosis occurred earlier for patients who switched antipsychotics than for those who did not, suggesting that switching is associated with an increased risk of relapse in patients with SZ, BD, and MDD. This may be attributed to more-severely ill patients being less responsive than those with less-severe illness, which, in turn, may require more episodes of switching.