In recent years, a clinical trial design called randomized phase II is often applied in cancer clinical trials. Among them, some trials are planned based on clinical trial designs, which are different from conventional one. As those conventionally used, there is a Simon's selection design, randomized screening design with concurrent control, and dose-finding randomized design. However, recently, many clinical trials corresponding to more complicated purposes different from these are often planned and conducted. Inherently, it is essential that trial designs that match the characteristics of the therapies or drugs to be developed are required, especially in phase II trial. Unlike phase I or phase III trials, which often remain in conservative designs, it is also natural that different phase II trials have different objectives. There should be various design options according to each trial objective. Considering the current applications, the randomized phase II design is a collection of heterogeneous designs. Researchers are required to refine properly in choosing an appropriate design for their clinical trials. Furthermore, depending on the needs of researchers, it is also necessary for biostatistician to develop a new design according to the present and coming era of medicine, for example, the era of molecular targeted therapeutic drugs and immune-oncology.