Objective: We used a large, employer-sponsored, administrative claims database of enrollees using second generation antipsychotics (SGAs) to address two study objectives: 1) describe patterns of SGA utilization by metabolic profile (MP) and polytherapy, and 2) examine the influences of MP and SGA polytherapy on hospitalization. Methods: Using descriptive and logistic regression analyses, we examined patterns of SGA polytherapy by MP, focusing on SGA use patterns and their association with hospitalization. MP status was categorized as low (aripiprazole, ziprasidone), moderate (risperidone, quetiapine), and high (olanzapine). Polytherapy of two or more SGAs was defined as long-term (i.e., augmentation for =90 days) and short-term (i.e., switching for <90 days). We controlled for sociodemographics, payor source, comorbidities, concurrent clozapine and typical use, and drug burden. Results: On average, individuals received 10.2 SGA scripts annually; 23.8% used low, 70.2% used moderate, and 22.8% used high MP SGAs. Second generation antipsychotics polytherapy occurred in 9.6% of all SGA users (7.3% in low MP, 8.4% in moderate MP, and 7.3% in high MP SGA users). Switching accounted for 77.4% of all polytherapy. Relative to high MP users, moderate MP users were slightly more likely to be hospitalized (moderate MP OR=1.2 [p<0.0001]). Low MP use was not significantly different from high MP use with respect to hospitalization. Relative to monotherapy SGA users, polytherapy users had hospitalization odds of 1.9 (p<0.0001). Other positive significant covariates included diagnoses of schizophrenia, bipolar disorder, depression, substance use disorders, obesity, increasing drug burden and atypical antipsychotic exposure. Clozapine use was negatively associated with all hospitalizations. Similar findings were found for psychiatric hospitalizations. Conclusion: Polytherapy and moderate MP SGAs are associated with higher risk of hospitalization compared to high MP SGAs and monotherapy use. These findings suggest the need for prudent selection of SGAs taking into account patients' comorbidities and pill burden.