In the last 5 years urologists have witnessed an explosion of articles that purport to predict outcomes of treatment for localized prostate cancer. We have seen articles that have found that preoperative prostate specific antigen (PSA), Gleason grade, percent high grade disease, tumor volume, clinical tumor stage, pathological tumor stage, age, perineural invasion, microvessel density and a whole slew of molecular markers are predictive of outcome. The outcomes used in these analyses have been final pathological stage or actual treatment results, that is PSA-free survival. Rarely has anyone reported the benefit of these biological markers in predicting actual patient survival. In this issue of The Journal of Urology we have 4 more articles that add further information to this large and confusing body of knowledge. Many academic institutions have used whole mount prostatectomy specimens to ensure complete tissue sampling and reconstruction to determine prostate cancer volume. A theoretical benefit of this technique is that more accurate tissue sampling will lead to greater detection of extraprostatic extension and surgical margin involvement. Hollenbeck et al (page 1583) demonstrate that for patients with intermediate grade disease and preoperative PSA less than 40 ng./ml. partial sampling of the prostate yields pathological staging results that are similar to those from whole mount sections. While it is possible that a small difference may exist in the precision of these 2 techniques or that there may be differences in high grade disease, it would appear that these differences are small and that, at least for routine clinical purposes, there is no significant advantage to the whole mount technique. Fujikawa et al (page 1587) have also added another tool to the prediction of pathological stage based on clinical parameters known preoperatively. They correlated mean nuclear volume determined with a computer imaging tool to pathological outcome. This technique may be analogous to flow cytometry or cellular cytology. Mean nuclear volume proved more powerful than Gleason grade for predicting pathological outcome and may be less physician dependent. Blute et al (page 1591) report on an analysis of 2,475 patients who underwent radical prostatectomy at the Mayo Clinic in which they used preoperative PSA, Gleason score and clinical stage to predict final pathological outcome. With minor differences, they found that the Partin tables reported originally in 1993 and updated in 1997 remain predictive of outcome.1 This finding should not be surprising to most readers as the updated validation of the Partin tables in 1997 was based on a large patient cohort from 3 different, highly reputable institutions. Nevertheless, this report once again validates that these 3 factors can be used to predict pathological outcome with reasonable accuracy. However, use of these tables is limited by the fact that correlation between pathological outcome and actual disease-free outcomes is, at best, loose. While these authors report that the Partin tables can be used to predict margin status, for example, they reported that 19% of positive margins were iatrogenic.2 While the 5-year biological disease-free survival of patients with negative margins is 86%, survival of those with a positive margin is only slightly worse at 75%.3