The diagnosis of uterine smooth muscle neoplasms by light microscopy is difficult. Multiple classification schemes have been proposed based on mitotic rate, nuclear atypia, and the presence or absence of necrosis. None of these classification systems has been entirely successful. This study was undertaken to evaluate the use of selected immunohistochemical and histochemical markers in differentiating these tumors, in addition to accepted morphologic criteria. Ten cases of each of the following: leiomyosarcomas (LMS), atypical leiomyomas (AL), cellular leiomyomas (CL) and usual leiomyomas (UL), were classically evaluated for histological diagnosis and were stained for Ki-67 (MIB-1), bcl- 2 and p53 using monoclonal antibodies and the avidin-biotin peroxidase method, and argyrophilic nucleolar organizer region (AgNORs). The number of stained cells was counted in the most positively stained region in a 4 mm2 square cover glass mounted on each slide. The mean value was calculated for each group of tumors. The data for Ki-67 (MIB- 1), bcl-2, p53 and AgNOR staining respectively, were significantly higher in LMS by comparison to UL, CL or AL. Because many singular cases had superimposed data being difficult to diagnose, a new scoring system for pathological evaluation was created. The results obtained by this scoring system suggest that immunohistochemical markers Ki-67 (MIB-1), bcl-2, p53 together with the AgNOR staining could be useful, by the scoring system, as an adjunct to the current accepted morphologic criteria in differentiating smooth muscle tumors of the uterus.