The Geriatric Social Work Competency Scale II is the leading measure of students' competency with older adults. Although the 40-item instrument has demonstrated strong internal consistency and congruent validity, the hypothesized four-factor structure has not been tested. We used confirmatory factor analysis with a sample of social work students (N=667) to test the fit of the hypothesized model and explore respecification. The original model demonstrated strong internal consistency and a roughly acceptable fit. However, high interfactor correlations and items with significant cross-loadings led to the development of a 36-item version. The reduced version made small but consistent improvements in model fit and discrimination among factors while retaining strong internal consistency. Implications for future research and scale usage are discussed.