Assess weight variations in Brazilian users of the 20 microg/day levonorgestrel-releasing intrauterine system (LNG-IUS, Mirena) for comparison with a cohort of users of the TCu 380A intrauterine device (IUD) and a cohort of users of the injectable contraceptive, depot-medroxyprogesterone acetate (DMPA) over a period of five years. A total of 163 users of the LNG-IUS, who had the device inserted in 1998, were admitted to the study. At the time of enrollment, each woman was matched by weight (+/- 1 kg) and age (+/- 1 year) to a woman using a TCu 380A IUD and to another woman using DMPA. All women were followed up for a period of five years. Each year weight was measured and the body mass index (BMI, kg/m2) was calculated. Age of the LNG-IUS users was 27.0 +/- 6.7 years (mean +/- SD), while age of the copper IUD users was 28.0 +/- 6.6 years and that of DMPA users was 26.9 +/- 6.5 years. Weight recorded at the onset of the study was 62.9 +/- 0.8 kg, 62.8 +/- 0.8 kg, and 62.5 +/- 0.9 kg (Mean +/- SD) for users of the LNG-IUS, the copper IUD, and DMPA, respectively. The BMI of women was initially 25.0 (+/- 0.3), 26.4 (+/- 0.3), and 25.5 (+/- 0.4), (Mean +/- SD), for users of the LNG-IUS, copper IUD, and DMPA, respectively. Weight increases of 3.1 kg, 4.9 kg and 8.2 kg were observed at the end of the fifth year among users of the LNG-IUS, copper IUD, and DMPA, respectively (p = 0.009). Increase of the BMI was also observed among all groups (final BMI was 26.3 +/- 0.7, 28.5 +/- 0.8 and 28.7 +/- 1.3 for users of the LNG-IUS, copper IUD and DMPA, respectively). Multivariate analysis showed that there was a significant association between weight increase and length of use of the contraceptive method and of the DMPA. The use of a LNG-IUS during five years caused no significant weight increase and the difference in weight was of the same magnitude as that of copper IUD users.5232-6