RT-PCR, if targeted against a prostate-specific marker, can be a highly specific and sensitive assay to detect the presence of occult prostate cancer cells at sites far distant from the primary tumor. The low false-positive rate (0.8%) observed when combining most published studies and the high rate of detection of prostate cells in metastatic patients (88%) found in well-defined clinical studies address the basic requirements of a clinically effective diagnostic modality. Additionally, all reports indicate that RT-PCR positivity for PSA or PSMA increases with increased stage of prostate cancer, suggesting that RT-PCR assays may have value in staging the patient presumed to have clinically localized disease. Indeed, data from two institutions have shown the unique contribution that RT-PCR technology might lend to this most vexing problem. With the suggestive data already published, we remain optimistic that molecular staging will become a reality in the future. After all, the variability in techniques of RT-PCR (as well as differences in targets and samples assayed) used by the various laboratories now investigating this method could themselves be responsible for many of the differences reported. There is no question that standardization among laboratories is essential before clinical utility of molecular staging can be proved. Indeed, this endeavor represents the major aim of the RT-PCR consortium in the United States. Nevertheless, at our own medical center, we are struck by our finding that if a preoperative RT-PCR for PSA is negative, the patient can be assured of a successful operative outcome in 88% of cases. We also are impressed by the data showing that if the serum PSA level is greater than 10 ng/mL and the RT-PCR assay is positive, 90% of patients undergoing radical surgery have adverse pathology reports (and, in fact, have already experience operative failure in nearly half the cases). There is increasing consensus that RT-PCR assays afford prognostic information that is also unique. Three institutions have similar data comparing operative or preoperative RT-PCR strategies (employing all three potential sample sources) with treatment outcome. This may suggest the validity of RT-PCR as a staging modality or, alternatively, suggest that a truly metastatic phenotype is identified in patients with circulating PSA-synthesizing cells, or in patients harboring PSA-synthesizing cells in node or marrow tissues. In either case, RT-PCR appears already to provide information not available before the use of this technology. In fact, in the only series comparing RT-PCR with other, more established predictors (PSA, Gleason score) in a multivariate analysis, RT-PCR status provided the best prognostic information. Prostate manipulation does, in fact, appear to affect the RT-PCR assay. Although digital rectal examination and cystoscopy do not seem sufficiently invasive to release prostate cells into the circulation, a small number of individuals undergoing transrectal ultrasound biopsy will experience this phenomenon. This observation provides a cautionary note relative to timing of sample collection for RT-PCR assays. Furthermore, a significant fraction of patients undergoing prostate manipulation during radical surgery may experience shedding of prostate cells into the operative field and release of some into the peripheral circulation. This has not been universally observed and, even if true, may not share the significance of spontaneous RT-PCR positivity (before prostatic manipulation). In the one instance, a potentially metastatic phenotype is responsible for RT-PCR positivity, whereas in the other instance, PSA or PSMA synthesizing cells may have no metastatic potential at all. Further studies on the biologic half-life of such cells are required before any clinical judgement can be made regarding this phenomenon. In summary, there appears to be consensus that RT-PCR technology can differentiate controls from patients with metas