Objective. To empirically determine if a substance abuse comorbidity is related to longer inpatient stays for patients diagnosed with schizophrenia and affective psychoses. Study Design. A cross-sectional analysis of patients in three states: Maryland, California, and Arizona. Using multivariate techniques, we control for the effects of patient severity, insurance, and hospital characteristics on length of stay. Data Collection. We used a patient-level and state-specific hospital discharge database merged with hospital characteristics from the American Hospital Association. Principal Findings. The conventional wisdom is that patients with a substance abuse comorbidity have longer inpatient stays than similar patients without a substance abuse comorbidity. We did not observe this trend. We found wide variation in length of stay by state and patients' health insurance. In some cases, length of stay was statistically shorter for patients with a substance abuse comorbidity. Conclusions. Our research demonstrates that a substance abuse comorbidity does not necessarily equate to longer inpatient stays and that previous studies may overstate the impact of substance abuse on inpatient care utilization. In addition, we find that a relationship between substance abuse and length of stay may not be generalizable across states, diagnoses, hospitals, and/or insurance types. Future studies of the impact of substance abuse on inpatient stays need to control for these important covariates. Key Words. Length of stay, inpatient stay, substance abuse comorbidity, schizophrenia, affective psychoses