The extent of tumor in prostatectomy specimens was determined by a grid method in 117 patients with prostatic adenocarcinoma. A plastic strip or ruler with squares of 3.0 mm was used, and the ratio of squares overlying carcinoma to the total number of squares overlying prostate tissue was calculated. This grid ratio, which represents an estimate of the percentage of the prostate involved by tumor, was a significant prognosticator closely tied to the likelihood of tumor progression and to survival time, as assessed by logistic regression analysis and a proportional hazard model. The grid ratio was better than histologic grade in predicting tumor progression and patient survival; also, the ratio was more objective than histologic grade as judged by interobserver agreement values. Only slight improvement in prognostication was obtained with concurrent use of both extent and grade. The grid ratio method was slightly better in predicting tumor progression and patient survival than a second method of assessing the percentage of prostatic tissue involved by tumor, the pathologist's percentage estimate. These results indicate that it is important to quantitate tumor extent within prostatectomy specimens; such quantitation need not require step-sectioning of the entire prostate and an expensive and time-consuming method such as computerized morphometrics but rather may be performed by a simple estimate of the percentage of the prostate involved by tumor. Reporting of histologic grade and tumor extent in the prostate gland is recommended as both appear to be important in identifying those patients at risk for a poor outcome after prostatectomy for prostatic carcinoma.