The cell proliferation activity of sixteen childhood primitive neuroectodermal tumors (PNETs) was observed immunocytochemically, to determine the cell kinetics and cell proliferation activity of these relatively undifferentiated, malignant brain tumors. Two mouse anti-human monoclonal antibodies (MoABs) were employed for the detection of the nuclear antigen (Ki-67) present during proliferation in frozen sections and proliferating cell nuclear antigen (PCNA) expression in formalin fixed, paraffin embedded tissue sections. A sensitive four step, indirect, streptavidin-biotin, alkaline phosphatase (AP) conjugated, immunocytochemical experimental technique, was used. The anti-Ki-67 MoAB which binds to a special nuclear antigen, identified its expression only in proliferating cells. This nuclear antigen was detectable during the whole mitotic cycle of all malignant and fast proliferating cells, only absent between and during phases G0 (resting phase) and the first gap phase (G1). The mean labeling index (MLI) was defined as the percentage of Ki-67 and PCNA antigen positive cells of the total number counted. The MLI for the PNETs ranged between 1.4% and 11.6%, with a mean MLI of 6.2% for Ki-67; and between 3.2% and 16.8%, with a mean of 9.74% for PCNA. All observed PNETs demonstrated heterogeneous nuclear stainings, but the highest MLIs were found among the poorly differentiated classic medulloblastomas (over 30% for the Ki-67 antigen and 46% for PCNA). MLIs were low in 5/13 PNETs (under 4% for antigen Ki-67 and 9.4% for PCNA) and in this group we defined our lowest (1.4%) MLI, suggesting the presence of in vivo neuritogenesis of these undifferentiated, embryonal tumors. MLIs in 6/13 PNETs were intermediate in magnitude. The MLIs were higher in two PNETs with clear cellular differentiation towards an ependymal (10.4%) and melanocytic (8.8%) direction. Formation of astrocytes within the tumor mass did not affect the intermediate character of the MLI (7.8%). The prognosis of every intracranial tumor is obviously correlated with its proliferation activity and cell kinetics. The clinical significance of these parameters is great since they provide direct information concerning the growth characteristics of an intracranial tumor.