<p>Only few quantitative studies have investigated identity development in older age, partly due to a lack of validated instruments. In this study, we evaluated the psychometric properties of a short (15-item) version of the Dimensions of Identity Development Scale (DIDS) in a cross-sectional sample of older Swedish adults (<i>N</i> = 3795; 54.6% women, 45.4% men; mean age 67.1&#xa0;years) before, during, and after transition to retirement. Though the scale had initially been developed for an adolescent population, model fit of the factor structure was acceptable and internal reliabilities of each dimension were high. Latent mean comparisons suggested that the subdimensions differ in stability across groups of participants in different stages of the retirement transition. While commitment was highest in bridge employees, all dimensions but exploration in depth were lowest in participants after retirement. Exploration in depth was the only dimension that was independent of stage in the retirement transition. Depressive symptoms correlated positively with ruminative exploration, but negatively with commitment making and identification with commitment. As such, these dimensions should be considered as possible risk/protective factors regarding mental health adversities in older adults who face disruptive life events. We discuss potential issues regarding the construct and content validity of the DIDS. A modification of the scale or the development of a new instrument based on a recent framework of identity integration could contribute to new and meaningful insights into identity development and its interplay with mental health across the lifespan. </p>