Current and recent users of combined oral contraceptives (COCs) have an increased risk of breast cancer, an effect which has not been consistently shown to be related to duration of use and which disappears within a few years of stopping. Current and recent users of COCs also have an increased risk of invasive cervical cancer, an effect which strengthens with prolonged use and which wears off after stopping. The risk of both ovarian and endometrial cancer is reduced in current users of COCs, a beneficial effect that increases with duration of use and persists for many years after stopping. Previous evidence that COC use may reduce the risk of colorectal cancer has not been confirmed in a number of more recent studies. Sparse evidence suggests that prolonged use of COCs increases the risk of hepatocellular carcinoma in women living in areas where the prevalence of hepatitis B virus is low. Lengthy follow-up of large cohort studies shows that the lifetime risk of any cancer in ever users of COCs is similar to that of never users. There have been comparatively few studies examining the cancer effects of progestogen-only contraceptives, especially in women who have used only these products. At the moment, it is prudent to assume these contraceptives have a similar pattern of breast and cervical cancer risk as COCs. Limited data indicates that users of non-hormonal intra-uterine devices may have a reduced risk of cervical cancer (any device), and endometrial cancer (for inert but not copper devices). Current evidence does not strongly suggest a changed risk of breast, cervical, endometrial or colorectal cancer among women who chose to be sterilised although they may be protected against ovarian cancer. Evidence linking vasectomy with an increased risk of prostate cancer is inconsistent and likely to be biased; an important association with testicular cancer has not emerged in the few studies conducted so far.