Posttherapy changes in prostate-specific antigen (PSA) have been used as a primary endpoint for patients with prostate cancer in an effort to rapidly identify promising therapeutic approaches. Using this method, agents that do not produce a defined decline in PSA on multiple determinations for a specified period are not to be pursued further. Multiple investigators have used PSA as an outcome measure, but the results are variable and often difficult to interpret. These inconsistencies are due to the lack of a consensus on criteria for PSA declines and the inappropriate use of PSA as an and point when it is not clinically applicable. Phase III studies are needed to validate any surrogate variable, including posttherapy declines in PSA; until such trials have been performed, investigators need to define and report all parameters of outcome independently.