BACKGROUND: Nicotine dependence is common in schizophrenia patients but rare in patients with obsessive-compulsive disorder. Little is known, however, about smoking in schizophrenia patients with obsessive-compulsive symptoms (OCS) especially in a developing country, such as Egypt, that has the highest rate of tobacco consumption in the Middle East and North AfricOBJECTIVES: To test a hypothesis that nicotine dependence in schizophrenia patients with-OCS is lower than in those without-OCS. RESULTS: Out of 87 consecutive schizophrenia patients attending a psychiatric out-patients clinic in Egypt, 34 patients (39%) had significant OCS, and a smoking rate (85.3%) not significantly lower than that of patients without-OCS (90.6%). Patients were reclassified by their Fagerström's scores into 3 groups of nicotine dependence: mild (37 patients), high (21 patients) and very high (29 patients). There were no differences between these groups in terms of YBOCS scores (F=0.324; p=0.724). When comparing PANSS scores of patients, with and without-OCS, those of the former group were higher on the positive symptoms (mean=24.2 versus 20.8; p=0.002), and anxiety/depression (mean=10.8 versus 10.1; p=0.03) but lower on the negative symptoms (mean=15.0 versus 19.4; p=0.000), disorganized thoughts (mean=14.5 versus 17.5; p=0.002), and uncontrolled hostility/excitement) (mean=7.5 versus 9.5; p=0.002). However, there was no significant between-group difference in the total PANSS scores. CONCLUSIONS: The high and equal rates of smoking in patients with-, and without-OCS, and the lack of difference between YBOCS scores of mild, high and very high nicotine dependence groups undermined our initial hypothesis that nicotine dependence in schizophrenia patients with-OCS is lower than in those without them. Nevertheless, the finding of a distinctive PANSS scores, may support a suggestion that OCS in schizophrenia represent a distinct subtype or dimension.