The purpose of this study was to examine the determinants for early discontinuation among IUCD users in a rural district of northern India. Multivariate analysis indicated several significant predictors of early discontinuation of IUCD use. The risk of discontinuation increased more than two times in the presence of factors such as more than usual amount of menstrual flow before insertion and intermenstrual bleeding after insertion. Similarly, residence or alternatively those coming from villages without health centres were nearly two times more likely to discontinue. Menstrual disturbance increased the risk nearly three-fold. If only those women who report normal menstrual cycles had IUCDs inserted, it is very likely that the high discontinuation rate due to menstrual disturbance could be reduced. Knowledge about the IUCD and its effectiveness decreased the risk of early discontinuation but was not a statistically significant factor in multivariate analysis. Validation of the developed model was done by using the bootstrapping method and the model was found to be 18% noisy. These findings may help family planning providers in counselling and practice.