DSM-III-R recently introduced early-late and primary-secondary subtypes of dysthymia. The present study explored the validity of the DSM-III-R early-late onset distinction by comparing early- and late-onset primary dysthymics on demographic, clinical and familial variables and short-term outcome. Compared to the late-onset dysthymics, the early-onset group had higher lifetime rates of superimposed major depressive episodes and anxiety disorders, had sought treatment significantly more frequently, had a higher rate of major affective disorders in first-degree relatives, and exhibited higher levels of depression throughout the course of a 6-month follow-up study. These data provide preliminary support for the validity of the DSM-III-R early-late onset distinction in dysthymia.