A randomized prospective trial of three copper IUDs, Nova-T, MLCu375 and MLCu250, including 200 of each, is presented. Insertion was done at the hospital outpatient clinic on normally menstruating women and on women in puerperio. Follow-up was scheduled after 12, 24 and 36 months. Pregnancy rates were low for all 3 models. Pearl indices after 3 years were 0.5, 0.9 and 0.8 for Nova-T, MLCu375 and MLCu250 respectively (NS). Abnormal bleeding and/or pain was the most frequent termination cause. Minor differences in the termination rates because of abnormal bleeding and/or pain were found and are discussed. The continuation rates based on all medically relevant IUD removals were 74%, 73% and 81% after 3 years for Nova-T, MLCu375 and MLCu250 respectively. No important difference in clinical performance between the three copper IUDs could be demonstrated.