Context: The issue of contraceptive use, its determinants and availability of varieties are presently a concern to the government of Lesotho. This is so because they are very important components of any successful family planning programs. Methods: The study uses the Lesotho Demographic and Health Survey (LDHS), 2004 data set to examine the role and pattern of three groups of variables (individual, fertility and contextual variables) in determining contraceptive use among women aged 15-49 years in Lesotho. Based on the Cost-Benefit theoretical model, analyses are done at three levels. These are the univariate, bivariate and multivariate levels. The logistic regression technique is used at the multivariate level. Results: All background characteristics at the bivariate level are found to have a strong significant association and hence a relationship with the use of contraceptive. This however is with the exception of sex of household head, which is only significant when controlling for marital status and age. At the multivariate level, women educational level, partner’s educational attainment and marital status have the highest odd ratio of contraceptive use and hence, the highest power of critical predictions. They are also highly significant in each model. The stepwise regression shows that addition of new group of variables (fertility and contextual variables), brought about changes in ratio values although very small. It also maintained the same pattern of predictions with the other models to the use of contraceptive. Conclusion: This study concludes that factors found to be associated with contraceptive use should be considered by the Lesotho government in its family planning programmes, especially those concerning contraceptives as this will bring about increase in use and hence, increase in contraceptive prevalence in the country.