The objectives of this study were to evaluate performance and causes of discontinuation of the Copper T380A IUD by users at the Family Planning Service at Hospital Arzobispo Loayza in Lima, Peru, during a period of three years. The study included 3167 acceptors of the CuT380A attending the service during 1992-1994. IUDs were inserted during interval timing. Follow-up was at 1,12,24 and 36 months after insertion. The following events were recorded: number of pregnancies, expulsions and all causes of discontinuation. The cumulative rates per 100 woman-years using the life-table method was calculated. Also calculated were the relative risks for expulsion and for pregnancy. At the end of the study, 361 women had discontinued the method for various reasons, whereas 1667 women continued using the method. The lost-to-follow-up proportion increased over time from 35.9 per 100 woman-years for the first year to 38.2 for the third year. The cumulative discontinuation rate over three years was 22.6±1.3 (cumulative rate±standard error) per 100 woman-years. The cumulative pregnancy rate for three years was 1.2±0.4 per 100 woman-years, whereas the cumulative rate of expulsion was 4.9±0.4 for the first, 6.4 for the second and 6.8 for the third year. The main cause of discontinuation during the first year of use was expulsion (4.9 per 100 woman-years) followed by personal reasons (2.1 per 100 woman-years). At the end of the third year, the main cause was personal reasons (11.4) and the second cause was expulsion (6.8). A higher probability of expulsion, pregnancy and dis-continuation for bleeding and/or pain was associated with age less than 20 years. In conclusion, the effectiveness of the CuT380A IUD after three years of use was 98.8 per 100 woman-years, whereas continuation was 39.2 and loss to follow-up increased over time.