A cross-sectional sample of adults, ages 20-79, were administered an adjective rating scale instrument measuring multiple affective states, including items from the Profile of Mood States (POMS) instrument. Confirmatory item factor analysis supported, for the most part, a priori assignments of items to scales based upon prior research, but revealed a few small, additional item factors that were cross-validated in a second sample. Items measuring different aspects of psychological distress, including anxiety and depressive affect, showed appreciable skew and kurtosis, with a substantial proportion of respondents indicating no perceived distress. Items measuring psychological well-being tended to show more normal response distributions. Tests of age-related invariance in item factor structure indicated that the unstandardized factor pattern weights (loadings) were not fully equivalent across two age groups, but showed that the same configuration of items loading on factors was supported. The scales perform well enough to justify continued use in older populations, but further research on the contributions of item distributions to age differences in factor loadings is needed.