Purpose: The present study investigated the association between nurse staffing and health outcomes among psychiatric inpatients in Korea by assessing National Health Insurance claims data. Methods: The dataset included 70,136 patients aged 19 years who were inpatients in psychiatric wards for at least two days in 2016 and treated for mental and behavioral disorders due to use of alcohol; schizophrenia, schizotypal and delusional disorders; and mood disorders across 453 hospitals. Nurse staffing levels were measured in three ways: registered nurse-to-inpatient ratio, registered nurse-to-adjusted inpatient ratio, and nursing staff-to-adjusted inpatient ratio. Patient outcomes included length of stay, readmission within 30 days, psychiatric emergency treatment, use of injected psycholeptics for chemical restraint, and hypnotics use. Relationships between nurse staffing levels and patient outcomes were analyzed considering both patient and system characteristics using multilevel modeling. Results: Multilevel analyses revealed that more inpatients per registered nurse, adjusted inpatients per registered nurse, and adjusted inpatients per nursing staff were associated with longer lengths of stay as well as a higher risk of readmission. More adjusted inpatients per registered nurse and adjusted inpatients per nursing staff were also associated with increased hypnotics use but a lower risk of psychiatric emergency treatment. Nurse staffing levels were not significantly associated with the use of injected psycholeptics for chemical restraint. Conclusion: Lower nurse staffing levels are associated with negative health outcomes of psychiatric inpatients. Policies for improving nurse staffing toward an optimal level should be enacted to facilitate better outcomes for psychiatric inpatients in Korea.