The aim of the present study was to examine the usefulness of the quantification of PC10-positive-cells and of Argyrophilic Nucleolar Organizer Regions (AgNORs) in gastric biopsies for the identification of gastric mucosal proliferative lesions. Fifty seven paraffin-embedded endoscopic biopsies were classified into four histologic groups: normal, inflammatory, dysplastic and neoplastic mucosa. The percentage of PC10-positive cells was determined by immunohistochemistry. The AgNOR parameters determined included the total number of all identifiable silver precipitations in the nucleus, the mean number of silver precipitations per cluster, and the presence of morphologically heterogenous silver precipitations. Group comparisons were performed using the Kruskall Wallis and Dunn non-parametric tests with a significance level of 5%. A discriminant analysis (followed by the jack-knife procedure) was performed using the three AgNOR parameters plus the percentage of PCNA-positive cells as the independent variables and histological groups as the dependent variable. All three AgNOR parameters, as well as the percentage of PCNA-stained nuclei, showed their highest values in the carcinoma group. However, no good differentiation among the four histologic groups was obtained using only one of these parameters, since there was always considerable overlap among them. By combining all the parameters in a linear discriminant analysis, we obtained a correct classification in 48 out of 57 cases. Within the classification errors there was only one false positive carcinoma, which was in fact a dysplasia and only one false negative carcinoma erroneously classified as dysplasia. The number of cells with heterogenous AgNORs was the most important parameter for the discriminant analysis. No correlation between PCNA values and the AgNOR parameters could be found, thus indicating that they do not represent the same phenomenon in the cell cycle. We concluded that the use of a combination of various proliferation parameters in a linear discriminant analysis may be helpful for differentiating gastric mucosal lesions. The peculiar AgNOR morphology is an important variable which should be taken in consideration in quantitative studies. PCNA and AgNORs seem to represent different physiological phenomena in the cell cycle.