OBJECTIVE: This study attempted to document a hypothesized increase in the frequency of the diagnosis of obsessive-compulsive disorder at a large psychiatric teaching hospital and to investigate correlates of this trend. METHOD: The annual rates of psychiatric discharge diagnoses at the hospital from 1969 to 1990 were reviewed, and the frequency of the diagnosis of obsessive-compulsive disorder was compared with that of paranoid disorders. Correlations were also done on these diagnostic rates and the rates of reports in the literature in the same years on each of these types of disorders and their treatment. RESULTS: The frequency of the diagnosis of obsessive-compulsive disorder, but not paranoid disorders, increased markedly during the 1980s. This increase was associated strongly and selectively with increases in publications about that disorder, particularly reports on drug and behavior therapy. CONCLUSIONS: There has been a large recent increase in the rate of diagnosis of obsessive-compulsive disorder, evidently associated with advances in the study and treatment of the disorder. The observations suggest the influence of a treatment-oriented diagnostic bias in which clinicians may more readily consider and diagnose a condition for which an innovative or effective treatment is available.