Like all drugs, combined oral contraceptives (COCs) have side effects that may be harmful or beneficial. During the last 20 years their adverse effects have been fully reported, but their benefits have been largely ignored. Most of the benefits of COCs result from the suppression of ovulation. This means that the advantages they confer are not dose-dependent, provided that ovarian activity is effectively suppressed. The most important health benefit of COCs worldwide is the effective prevention of pregnancy, which carries high risks in developing countries and has a mortality as high as 1 in 150 in Africa. The risk of ectopic pregnancy is reduced by 90% in COC-users compared with women using no contraception. The COC prevents the repeated proliferation of ovarian and endometrial tissue that takes place in the menstrual cycle, and it is therefore not surprising that it reduces the risk of ovarian and endometrial malignancy. What is surprising is that a relative risk of 0.6 for these cancers can be detected after only 12 months or less of COC use, and persists for at least 15 years after the COC is stopped. The COC reduces the incidence of benign breast disease, though not the types of disease linked with breast cancer. It considerably reduces the incidence of benign ovarian cysts, and this has been calculated to avoid 28 operations for functional ovarian cysts per 100,000 pill users every year. The risk of uterine fibroids is reduced by 17% with every five years of COC use. By thickening the cervical mucus, the COC reduces the risk of pelvic inflammatory disease by about 50%. By inhibiting growth and development of the endometrium it reduces the incidence of menorrhagia and consequently iron-deficiency anaemia, and it produces a 40% reduction in the frequency of dysmenorrhoea. Unlike the benefits of the COC, its risks appear to be to some extent dose-dependent. The first serious risk to be discovered was a three- to six-fold increase in venous thromboembolism, which is probably an oestrogen effect and disappears quickly when the COC is stopped. The COC doubles the risk of haemorrhagic stroke, this risk is related to smoking and hypertension, unlike the increased risk of thrombotic stroke. The risk of myocardial infarction is related to smoking and age, and COCs are contraindicated over the age of 35 in smokers though not necessarily in non-smokers. Much of this information, however, is based on studies involving older high-dose COCs. Risks may well be lower with modern COCs, but firm data are lacking.(ABSTRACT TRUNCATED AT 400 WORDS)