In cancer clinical studies, sequential implementation of "clinical trials" is mandatory in order to prove an effect of a proposed new treatment. Trials should be performed to prove safety in phase I and effectiveness in phase II trials. Then in phase III trial, comparison between the new and the "standard" treatment is performed. Many pitfalls and errors may occur when conducting trials, and also at the time of biostatistical analyses and their interpretation. Explanations of several such pitfalls in cancer clinical trials are attempted by the authors.