OBJECTIVE The utility of an antibody to proliferating cell nuclear antigen (PCNA), a growth-specific nuclear protein, was assessed as a prognostic variable for prostatic adenocarcinoma. Its expression was correlated with established prognostic indicators, including tumor grade, stage, prostatic-specific antigen (PSA), and percent of tumor in the gland at excision. METHODS Forty archival needle biopsies containing a minimum of four hundred tumor cells were analyzed. Immunoperoxidase staining of paraffin sections was performed for PCNA (PC10) after pretreatment in antigen retrieval solution. A proliferative index (PI) for each case was derived using image analysis with measurement of at least four hundred twenty-five nuclei. RESULTS PI values ranged from 2.4 to 31.3 percent. Mean PI values varied significantly (ANOVA, p = 0.005) among cases with dominant Gleason grade (DGG) of 3 (mean PI = 9.3%), 4 (mean PI = 13.7%), and 5 (mean PI = 18.8%). By t test, significant differences were noted for PI in cases with DGG 2 and 3 versus those with DGG 4 and 5 (p = 0.0065). PI for cases with DGG 3 versus 5 showed significant difference (p = 0.0017). Tumors of Gleason scores 5 to 7 differed significantly from those with scores 8 to 10 (p = 0.014). A statistical relationship for PI and PSA, clinical stage, and percent tumor at resection could not be established by linear regression. CONCLUSIONS These findings suggest that additional study of the PI, as determined by PCNA immunohistochemistry and image analysis, may be warranted to determine its usefulness as an adjunctive parameter in prostate adenocarcinoma. This technique may be particularly useful in needle biopsies where limited tumor may render assessment of grade difficult.