Social withdrawal is a symptom experienced by individuals with an array of mental health conditions, particularly those with schizophrenia and mood disorders. Assessments of social withdrawal are often lengthy and may not be routinely integrated within the comprehensive clinical assessment of the individual. This study utilized item response and classical test theory methods to derive a Social Withdrawal Scale (SWS) using items embedded within a routine clinical assessment, the RAI-Mental Health (RAI-MH). Using data from 60,571 inpatients in Ontario, Canada, a common factor analysis identified seven items from the RAI-MH that measure social withdrawal. A graded response model found that six items had acceptable discrimination parameters: lack of motivation, reduced interaction, decreased energy, flat affect, anhedonia, and loss of interest. Summing these items, the SWS was found to have strong internal consistency (Cronbach's alpha = 0.82) and showed a medium to large effect size (d = 0.77) from admission to discharge. Fewer individuals with high SWS scores participated in social activity or reported having a confidant compared to those with lower scores. Since the RAI-MH is available across clinical subgroups in several jurisdictions, the SWS is a useful tool for screening, clinical decision support, and evaluation.