Prostate glands from 52 patients with clinical stage B carcinoma were examined using two sampling techniques. After fixation and conization of the apical portions, each gland was serially sectioned with sections mounted whole on oversized glass slides and examined for pathologic features of prognostic importance. A second examination was subsequently conducted on the same tissue using only alternate sections. No differences in tumor type, grade, Gleason score, multiplicity, or capsular penetration were detected in 75% of cases. The discrepancies that did occur were most often minor variations in multiplicity and Gleason score. Of the 20 glands with capsular penetration observed with the serial sectioning method, 17 (85%) were detected using alternate sectioning. The surgical margin was involved in two of the three invasive foci that would have been missed. Although the topography is better displayed, the authors' examinations indicated no significant advantage to whole mount sections compared with sections mounted on standard-sized glass slides. Considering the most effective use of resources, as well as the current modalities available for patient monitoring, the results support the use of an alternate sectioning method for pathologic examination of specimens removed for clinically localized prostate cancer.