We examined 151 entirely submitted transurethral resection of the prostate (TURP) specimens that showed incidental prostate cancer. Fifty-one (34%) showed > 15% involvement of the specimen by tumor; these were not reviewed because submission of more tissue would not convert these cases from stage T1b (> or = 5% of tumor) to stage T1a ( < 5% of tumor). Sixty-six cases (44%) that were totally submitted in < or = nine cassettes also were excluded because most laboratories would totally submit these specimens from the outset. The remaining 34 cases (22%) had < or = 15% tumor and required 10 or more cassettes for total submission. The first eight slides were reviewed and percentage of tumor involvement and grade was calculated. The remaining slides were then reviewed to see if the overall tumor percentage or grade changed. Because in one case the tumor grade was significantly increased in the remaining slides and of the infrequency with which more than eight cassettes needs to be submitted, we recommended submission of all remaining tissue in stage T1a lesions. There is no need to submit additional tissue in stage T1b lesions because the percentage will not decrease with greater sampling.