Background: Epidemiologic studies of air pollution and mental health are limited. We examined ozone (O3), carbon monoxide (CO), nitrogen dioxide (NO2), and fine particles (PM2.5), and emergency room (ER) visits from all mental disorders, bipolar disorder, depression, schizophrenia, substance abuse, homicide/inflicted injury, and suicide/self-harm. We also evaluated effect modification by season, age, race/ethnicity, and gender.Methods: Air pollutant data, provided by the US EPA, and ER visit data from the California Office of Statewide Health and Development Planning, were merged from 2005 through 2013. Data analyses were conducted using two-stage quasi-Poisson regression models, adjusting for long-term trends, holidays, day of the week, and mean apparent temperature. Several lag periods were considered (same day up to average of 7 days and average of 30 days).Results: Per interquartile range, weekly mean O3 was associated with a 1.87% (95% confidence interval [CI]: 0.62, 3.15) increase in ER visits for depression (n = 162,451), a 1.43% (0.35, 2.51) increase for suicide/self-harm (n = 198,509), and a 2.83% (1.53, 4.15) increase for bipolar disorder (n = 158,906). O3 (30 day lag; n = 632, 787), CO (2 day lag; n = 229,138) and NO2 (2-day lag; n = 206,919) were associated with greater risks for homicides/self-inflicted injury (2.01% [1.01, 3.02]; 2.3% [1.3, 3.3]; and 2.6% [1.2, 4.0], respectively). Greater risk was found during the warm season for CO and NO2. For some outcomes, differential risk was observed by gender, age, and race/ethnicity. These associations generally remained robust in co-pollutant models. No associations were found between PM2.5 and any of the mental health outcomes considered.Conclusions: Our results suggest that short-term gaseous air pollution exposure increases the risk of homicides/inflicted injury (CO, NO2 and O3), and depression, suicide/self harm, and bipolar disorder (latter three outcomes for O3 only) in California.