Religiosity has been linked to mental health outcomes for decades. This study examined the potential relationship between religiosity and demographic and clinical variables in a sample of psychiatric inpatients. In total, 688 adults admitted to an acute psychiatric facility with a primary mood or psychotic disorder completed the Duke University Religion Index (DUREL). The DUREL measures religious activity in 3 domains: organizational religious activity (ORA), nonorganizational religious activity (NORA), and intrinsic (or subjective) religiosity (IR). We categorized scores into high and low religiosity. Bivariate analyses with χ and independent sample t tests were used to examine the association between the DUREL subscales and demographic, clinical, and outcome measures. A generalized linear model was used to identify predictors of suicidality, psychosis, and 30-day rehospitalization. Elevated religious activity was common in the inpatient sample, with 58% categorized as high IR, 43% as high NORA, and 36% as high ORA. For all 3 DUREL subscales, high religiosity scores were associated with significantly more psychosis (P<0.05) and significantly less suicidal ideation (P<0.001). High ORA (P=0.001) and high IR (P=0.01) were associated with significantly fewer suicide attempts. High ORA scores were also associated with an increased length of stay (P<0.05) and more frequent 30-day readmission rates (P=0.01). In the generalized linear model, predictors of lower levels of suicidality were high ORA, high IR, and a diagnosis of schizophrenia, schizoaffective disorder, or other psychotic disorder, whereas a diagnosis of depressive disorder was associated with greater suicidality. Predictors of psychosis were high IR and Hispanic ethnicity, whereas a diagnosis of depressive disorder was associated with lower rates of psychosis. Female inpatients were more likely than male inpatients to score high on the ORA (P<0.05), NORA (P<0.05), and IR (P<0.0001) subscales. In addition, a significant relationship was detected between age and high IR scores (P<0.005), with increasing age associated with higher IR scores. Although preliminary, these results suggest that a brief measure of religiosity may provide important information concerning clinical features and acute outcomes in patients hospitalized with serious mental illness.