Readmission after inpatient care for a psychiatric condition is associated with a range of adverse events including suicide and all-cause mortality. This study estimated 30-day readmission rates in a large cohort of inpatient psychiatric admissions in New York State and examined how these rates varied by patient, hospital, and service system characteristics. Data were obtained from Medicaid claims records, and clinician, hospital, and region data, for individuals with a diagnosis of any mental disorder admitted to psychiatric inpatient units in New York State from 2012 to 2013. Psychiatric readmission was defined as any unplanned inpatient stay with a mental health diagnosis with an admission date within 30 days of being discharged. Unadjusted and adjusted odds ratios of being readmitted within 30 days were estimated using logistic regression analyses. Over 15% of individuals discharged from inpatient units between 2012 and 2013 were readmitted within 30 days. Patients who were readmitted were more likely to be homeless, have a schizoaffective disorder or schizophrenia, and have medical comorbidity. Readmission rates varied in this cohort mainly because of individual-level characteristics. Homeless patients were at the highest risk of being readmitted after discharge.