This introduction to a symposium on the relationship between contraception and cancer presents some important preliminary considerations. Several factors acting independently are required to cause cancer, therefore a patient's risk factors should be evaluated before prescribing contraception. The statistics are difficult to interpret because dozens of different oral contraceptive pill formulations and doses have been used. Thousande of subjects would be required to prove a specific risk. Furthermore these risks should be judged in relation to others such as pregnancy, abortion, obesity, and smoking. Most studies available do not meet statistical criteria in either protocol or sufficient numbers. It is tempting to extrapolate from theoretical data, in this case one must indicate so, to prevent spread of rumor. Finally, constant surveillance of all contraception patients will aid in ascertaining cancer.