Combination injectable contraceptives provide a highly effective, reversible method of preventing pregnancy, and they do not require daily administration or use at the time of coitus. Although they are used in many countries, their acceptability could be limited by method characteristics, such as the need to obtain a monthly injection or bleeding pattern changes. To assess the contraceptive efficacy, bleeding patterns, discontinuation, user preferences, and side effects of combination injectable contraceptives. We searched computerized databases for randomized controlled trials of combination injectable contraceptives. Randomized controlled trials were eligible if they compared a combination injectable with any other contraceptive method (e.g., a second combination injectable contraceptive, progestin-only injectable contraceptive, other hormonal contraceptive or barrier method) or placebo. We limited the review to currently marketed combination injectable contraceptives. One author evaluated all titles and abstracts from the literature searches to determine their eligibility. Two authors independently extracted data from the eligible trials. Data on contraceptive efficacy, bleeding patterns, continuation, and side effects were entered and analyzed with RevMan. Combination injectable contraceptives include depot medroxyprogesterone acetate (DMPA) 25 mg plus estradiol cypionate (E(2)C) 5 mg, as well as norethisterone enanthate (NET-EN) 50 mg plus estradiol valerate (E(2)V) 5 mg. These contraceptives resulted in lower rates of early study discontinuation due to amenorrhea or other bleeding problems than progestin-only contraceptives. However, rates were higher for overall discontinuation and discontinuation due to other medical reasons. Acceptability results favored the combination injectable in one study and the progestin-only in another.Studies comparing two combination injectable contraceptives found that NET-EN 50 mg plus E(2)V 5 mg resulted in less overall discontinuation and less discontinuation due to amenorrhea or prolonged bleeding than DMPA 25 mg plus E(2)C 5 mg. However, these differences were not detected in all trials. The NET-EN plus E(2)V group also had more regular bleeding and fewer prolonged bleeding reference periods than the DMPA plus E(2)C group. The groups did not differ in their amenorrhea rates. While discontinuation rates can be viewed as a measure of method acceptability, the findings should be interpreted with caution since discontinuation depends on many factors. Future research should be directed toward interventions to improve the acceptability of combination injectable contraceptives, such as providing injections in settings more convenient than clinics, methods for women to administer their own injections, and counseling about possible bleeding pattern changes.