The success of cancer chemotherapy has increased greatly the potential areas for its inclusion in clinical trials and therefore has made the experimental design considerations more complex in these trials. New drugs still go through the three classic phases of clinical study beginning with clinical pharmacology (Phase I), efficacy screening (Phase 2) and role delineation (Phase 3). The Phase 2 and 3 trials now need to be considered within the overall therapeutic strategies which are required for each of the many diseases which are called cancer. The exigencies of a disease-oriented strategy which requires a blend of therapeutic modalities many times require a modification of what would be an ideal modality-oriented strategy geared solely to effectively testing a new agent. A new drug may need many Phase 2 and 3 trials in a variety of tumors before it can be considered adequately evaluated. Essential aspects of valid clinical trials include an adequate protocol which details the study and a design which is feasible for answering the question. Historical controls are advocated by many as a valid approach to clinical trials. Unless the numbers in the historical control are large and the prognostic variables well matches this approach should be viewed cautiously by the practicing physician.