BACKGROUND Behçet's disease is a vasculitis with diverse systemic manifestations usually involving the skin, eyes, joints and nerves. Clinical diagnosis is based on several signs, generally using the criteria established by the International Group for the Study of Behçet's disease, by O'Duffy, by Hamza, by Mason and Barnes and by the Japanese Committee. The purpose of this retrospective study was to assess the frequency of clinical manifestations in Behçet's disease and to apply the five diagnostic criteria scorings cited above to our patients to determine their sensitivity. PATIENTS AND METHODS Between January 1988 and December 1997, 110 cases of Behçet's disease were seen at our Dermatology Unit in Casablanca as inpatients or outpatients. Complete history and clinical data including results of the ophthalmology exam and skin tests were available for 70 patients (29 men and 41 women, mean age 31 years). RESULTS All 70 patients had buccal aphtosis and 88 p. 100 had genital aphtosis. Ocular signs were found in 36 p. 100 of the patients and the skin test was positive in 57 p. 100. According to the criteria of the International Group for the Study of Behçet's disease, 65 patients (93 p. 100) had Behçet's disease. For the Japanese Committee criteria there were 63 (90 p. 100), for the O'Duffy criteria 48 (68 p. 100), for Hamza criteria 45 (64 p. 100) and for Mason and Barnes criteria 35 (50 p. 100). DISCUSSION Our series is comparable to other Mediterranean series observed in dermatology units, but our patients had fewer systemic manifestations compared with another Moroccan series reported from an internal medicine ward. This could be due to recruitment bias and different medical education. For positive diagnosis of Behçet's disease, the criteria established by the International Group for the Study of Behçet's disease, the Japanese Committee, O'Duffy and Hamza enabled the diagnosis in the majority of the cases. Fifty percent of the patients had all the criteria described in these scores. These criteria would appear to be both overly descriptive, with the risk of confounding other disease states such as enterocolopathy, or underly descriptive, making it impossible to classify certain patients as having Behçet's disease.