The purpose of the present paper was to identify the rate of prevalence of impulse control disorders (ICD) in patients with obsessive-compulsive disorder (OCD) and to compare patients with OCD with and without ICD with regard to sociodemographic, clinical and prognostic characteristics. Forty-five patients with OCD were assessed by means of the Structured Clinical Interview for Diagnostic and Statistical Manual of Mental Disorders (4th edn, DSM-IV) plus additional modules for the assessment of ICD and examined using the Yale-Brown Obsessive-Compulsive Scale, the Clinical Global Impression, the Beck Depression Inventory, the Hamilton Depression Rating Scale, and the Global Assessment of Functioning. These patients were treated with serotonin re-uptake inhibitors (SRI) and followed for a variable period of time. Individuals with ICD (here defined as including not only the impulse control disorders not elsewhere classified of the DSM-IV, but also other disorders in which impulse control is a prominent feature such as alcohol and drug dependence, paraphilias and bulimia nervosa/binge eating disorder) were compared to those without ICD using the Mann-Whitney U-test and the Pearson's goodness of fit chi2 test. Sixteen patients with OCD (35.5%) displayed comorbid ICD. Patients with ICD were characterized by a significantly earlier age at OCD onset (P=0.04), a more insidious appearance of OCD symptoms (P=0.04), a higher rate of comorbid anxiety disorders (P=0.03), a greater number (P=0.02) and severity of compulsive symptoms (P=0.04), an increased rate of counting compulsions (P=0.02), and a higher number of required SRI trials (P=0.01). When OCD is found in association with ICD, the clinical picture is characterized by a greater severity of the obsessive-compulsive symptoms at presentation and by the requirement of a greater number of therapeutic attempts during follow up.