OBJECTIVE This report compares patients with the hypercoagulable state of polycythemia vera to patients with secondary polycythemia caused by tobacco use to determine whether the incidence of thromboembolic events is equivalent. METHODS The medical records of 146 patients with the diagnosis of polycythemia between 1977 to 1990 were reviewed. Patients with transient, relative, or stress polycythemia were excluded from this study as were patients with polycythemia as a result of cardiac or pulmonary anomalies or both. The diagnosis of polycythemia vera (n = 43) was verified by use of the guidelines of the Polycythemia Vera Study Group. The diagnosis of polycythemia caused by smoking (n = 27) was based on an elevated total red blood cell volume, decreased oxygen saturation on arterial blood gas measurement, evidence of chronic obstructive pulmonary disease, and elevated carboxyhemoglobin levels. RESULTS Twenty-six patients (60%) with polycythemia vera and 11 patients (41%) with smoker's polycythemia had at least one thromboembolic problem. No significant differences existed between the groups with regard to age, hematocrit, or number of cardiac and cerebrovascular events. Overall, patients with polycythemia vera had a greater number of thromboembolic events per patient (p < 0.05) and more peripheral arterial thromboemboli (p < 0.005) than did patients with polycythemia as a result of smoking (Fisher's Exact Test). CONCLUSIONS Thus the results of this study demonstrate that smokers' polycythemia does not represent a hypercoagulable state equivalent to that of polycythemia vera.