Gleason's score (GS) has been reported to be the most valuable prognostic factor in cases of prostate cancer. GS is solely dependent on the histological architecture of the prostate cancer, but, it seems doubtful that histological patterns are sufficient for evaluating the degree of malignancy of prostate cancer. We previously reported that the estimation of volume-weighted mean nuclear volume (MNV) might be a more useful prognosticator for prostate cancer than subjective histological grading. However, the previous study was conducted on patients treated in a single hospital, and the number of subjects was too small to draw a definitive conclusion. In this study, we analyzed a larger number of subjects at another institution using a blinded study design. A retrospective prognostic study of 195 patients with prostate cancer diagnosed between January 1966 and December 1988 at Kyoto University Hospital, and treated by conservative therapy, was conducted. Unbiased estimates of MNV were compared with the clinical stage and histological grading according to GS with regard to the prognostic value. Univariate analysis revealed that estimates of MNV, clinical stage, and GS all correlated significantly with disease-specific survival in cases of prostate cancer. Multivariate analysis of all cases also revealed that all of these factors were significant independent prognosticators of disease-specific survival. However, focusing on clinically localized cases (stages A, B, and C), multivariate analysis revealed that the estimation of MNV was the only powerful prognosticator of prostate cancer. This study indicates that the estimation of MNV is prognostically equal or superior to GS in cases of prostate cancer. We emphasized that the estimates of MNV is a more objective method for histological grading to predict the malignant potential of prostate cancer.