The discovery of multiple putative therapeutic targets and multiple putative agents for these targets in prostate cancer in the coming years poses significant challenges for clinical trial design. This is especially true for cytostatic agents that are not expected to lead to frank tumor shrinkage or declines in the PSA. The most promising agents will need to be identified early in their development. Since surrogate biologic markers are likely to play a critical role, the identification and validation of these markers is discussed. A number of non-traditional phase I and phase II clinical trial designs, including pre-operative dosing for assessing drug effect on a marker and the randomized discontinuation phase II design, are also discussed in detail. Use of such designs as well as surrogate marker validation will likely be required to efficiently choose appropriate agents for definitive study in the phase III setting.