The value of digital rectal examination, computerized tomography, magnetic resonance imaging, prostate-specific antigen, transrectal ultrasonography, and systematic-sextant biopsy in the identification of lymph node-positive patients before radical prostatectomy was analyzed in 103 men who had pelvic lymph node dissection, CT had a sensitivity of only 7% and a specificity of 96% in detecting lymph nodes, whereas magnetic resonance imaging had a sensitivity of 50% and a specificity of 100%. To evaluate the use of tumor volume in predicting lymph node metastasis, we counted the number of positive core biopsies and compared the results with the incidence of positive lymph nodes. If fewer than 5 positive core biopsies were considered negative for predicting lymph node metastasis, the sensitivity would be 67% (12 of 18), and the specificity 94% (50 of 53). To investigate tumor volume more precisely, we measured the extent of tumor volume in every biopsy as a percentage of the total biopsy core and added the percentage for the 6 biopsies. The lowest score was 10% (10% prostatic cancer in 1 of 6 cores), the highest score 580% (4 cores with 100% each and 2 with 90% each). The score was analyzed for sensitivity and specificity in predicting lymph node metastasis. If a score of 280% was used as a cutoff point, the sensitivity was 71% (10 of 14) and the specificity 91% (52 of 57). When we include the grading system by multiplying the percentage of tumor volume with tumor grade, the difference between the lymph node-positive state and lymph node-negative state becomes even more readily apparent.(ABSTRACT TRUNCATED AT 250 WORDS)