After nearly two decades of clinical use of oral contraceptives, clinical studies have shown a few rare adverse reactions, including jaundice, venous thrombosis and pulmonary embolism, ischaemic cerebrovascular effects, hypertension, amenorrhoea, breakthrough bleeding, anxiety, depression, weight gain, and changes in libido. Preclinical studies in rodents dosed with oral contraceptives did not and would not be expected to reveal side effects of such low incidence. A carcinogenic effect can be produced in certain strains of mice and rats when oestrogens, progestogens and combinations of these hormones are given in high doses throughout their life span. The susceptibility to tumour induction by hormonal contraceptives is not consistent in different strains of mice and rats, and if there is little or no predictive value in rodents it is difficult to see how they can provide useful information of the potential carcinogenicity of any of these compounds in women. Ideally, carcinogenic testing of oral contraceptives should be conducted in animal species which have a similar metabolic disposition of the drugs and similar feedback mechanisms which control ovulation and menstruation to that of women: subhuman primates and guinea pigs possess some of these characteristics.Many previous studies indicate that carcinogenic tests in rodents have not been carried out in a very satisfactory manner.