Phase III studies are the last step in evaluation of the new drugs or treatment modalities, and have one of the following purposes: 1) to determine the effect of a treatment relative to the natural history of the disease; 2) to determine whether a new treatment is more effective than a standard therapy; 3) to determine whether a new treatment is as effective as a standard therapy but is associated with less toxicity. In general, a phase III study is performed in a disease-oriented prospective randomized setting. The design of phase III study is described in this paper. In designing a trial, we must decide the endpoints which are appropriate for comparison of each treatment. The endpoints for phase III study include tumor shrinkage, disease-free survival, overall survival, toxicity and quality of life. In a phase III study, it is very important to determine how many patients should be studied (sample size). The sample size that will be needed for each endpoint depends on the expected difference between the standard and the experimental arms. Once the required number of patients is determined, the investigator must verify that there will be enough patients available to complete the trial. The most common cause of an uninterpretable result in these trials is the accrual of too few patients to make a definitive statement.