Objective: Evaluation of proliferative activity in diagnostics of diffuse large B cell lymphoma (DLBCL) is crucial on the one hand for specifying the diagnosis and on the other hand for estimation of chemosensitivity and eventually for determination of the outcome of the disease. Materials and Methods: We analyzed the archived materials of the cases, registered through 2004-2010 in the department of pathology of Academician N.Kipshidze central university clinic. Ten cases of centroblastic variant of DLBCL (I group), 20 cases of immunoblastic variant (II group) and 8 cases of anaplastic variant (III group) were chosen for the research. The slices of tissue were stained using H&E technology, immunohistochemistry (used markers: CD20, CD79, CD3, CD43, CD30, Bcl2, Bcl6, Ki-67). The parallel slices of each case were stained using AgNOR technology. We estimated the number of AgNOR positive cells and AgNOR positive grenules per nucleus. The numeric results were expressed as mean ± SD, and the 95% confidence intervals (CIs) of the means were calculated. Relationships between variables were analyzed using Pearson’s correlation analysis. Results: The result of the research showed: the evaluation of proliferative index (Ki-67) in all three variants of DLBCL split the cases in two subgroups: I - tumor with high proliferative activity (ki-67>10%) - 4 cases of centroblastic variant (40%), 12 cases of immunoblastic variant (60%) and 6 cases of anaplastic variant (75%); and II - tumor with low proliferative activity (Ki-67<10%) - 6 cases of centroblastic variant (60%), 8 cases of immunoblastic variant (40%) and 2 cases of anaplastic variant (25%). Using AgNOR technology revealed positive correlation between the number of AgNOR positive cells and high proliferative activity in I group and no correlation was found with low prolifrative activity. In second group the number of AgNOR positive cells, but not AgNOR positive granules, was in positive correlation with high proliferative activity. There was no significant correlation between AgNOR positive cells, AgNOR positive granules and proliferative activity (low/high) in III group. Conclusion: AgNOR technology can be used as an additional (auxiliary) diagnostic method for diagnostics of centroblastic and immunoblastic variants of DLBCL. And for evaluation criteria we should choose the number of AgNOR positive cells and not AgNOR positive granules.