(1) Intrauterine devices (IUDs) are placed in the uterine cavity with the objective of providing long-term contraception, mainly by preventing fertilisation. The best-known IUDs contain copper, but there is also an IUD delivering levonorgestrel, a progestin; (2) How effective are these devices, and what are their adverse effects? To answer these questions, we analysed the literature using the standard Prescrire methodology; (3) T-shaped copper IUDs, with a copper surface area of 380 mm2 on 3 arms, and the levonorgestrel-releasing device, have similar contraceptive efficacy as combined oral contraceptives that are used correctly. In contrast, IUDs are more effective than oral contraception used incorrectly; (4) Among IUD users, there are on average about 6 pregnancies per 1000 woman-years. There is less experience with the levonorgestrel IUD which seems to be at least as effective as copper IUDs; (5) The rare intrauterine pregnancies that occur in women using an IUD generally end in miscarriage. About 25% of these pregnancies end in a live birth if the device is left in place, compared to about 90% if the device is removed; (6) Ectopic pregnancies are rarer in IUD users than in women who do not use contraception. However, about one in 20 pregnancies that occur in women using an IUD is ectopic; (7) The IUD is expelled in about 5% to 10% of cases within 5 years, and expulsion recurs in about 30% of these women; (8) Problems such as difficult insertion, pain, bleeding and syncope are reported in less than 1.5% of cases overall; (9) Uterine perforation during insertion is rare, occurring in 0.6 to 16 cases per 1000 insertions, regardless of the type of IUD. The risk of perforation is higher when the IUD is inserted less than 4 to 6 weeks after delivery or elective abortion; (10) During the first 3 months after insertion, the risk of pelvic infection is slightly higher than in the general population, especially in women with pre-existing asymptomatic Chlamydia trachomatis infection. There are about 6 pelvic infections per 1000 woman-years of IUD use. Routine antibiotic prophylaxis is unnecessary. The interview and physical examination may lead to diagnosis of C. trachomatis infection or other sexually transmitted infections. In these cases, treatment may be needed before IUD insertion. Women must be warned that IUDs do not protect them from sexually transmitted diseases; (11) Menstrual bleeding is often heavier in women with cooper IUDs than in women who do not use IUDs, and may be associated with menstrual pain; (12) The levonorgestrel IUD is associated with a marked reduction in menstrual blood loss and irregular bleeding; amenorrhoea occurs in 35% of women after 2 years of use. The levonorgestrel IUD also has hormonal adverse effects such as headache, acne, breast tension and functional ovarian cysts; (13) IUDs can safely be used in breastfeeding women, immediately after a pregnancy, in cases of diabetes or HIV infection, during nonsteroidal antiinflammatory drug therapy, and after an ectopic pregnancy. The only problems occurring in women who have never had children are pain during insertion and more frequent expulsions; (14) A copper IUD is a first-line contraceptive method for women with a history of deep venous thrombosis, pulmonary embolism, or coronary events; (15) It is better to postpone IUD insertion when the woman has a genital tract infection or unexplained vaginal bleeding; (16) IUD insertion is an effective alternative to "morning-after" hormonal contraception.