This paper uses data from the , 1993 and 1998 Surveys to examine trends and determinants of contraceptive method choice. The analysis, based on two-level multinomial regression models, shows that, across years, use of modern contraceptive methods, especially long-term methods is higher in the urban than rural areas, while the pattern is reversed for traditional methods. Use of barrier methods among unmarried women is steadily rising, but the levels remain disappointingly low, particularly in view of the HIV/AIDS epidemic in Kenya. One striking result from this analysis is the dramatic rise in the use of injectables. Of particular program relevance is the notably higher levels of injectables use among rural women, women whose partners disapprove of family planning, uneducated women and those less exposed to family planning media messages, compared to their counterparts with better service accessibility and family planning information exposure. There is substantial evidence from existing literature that broadening the choice of contraceptive methods increases overall contraceptive prevalence (Freedman and Berelson, 1976;Phillips et al., 1988; Ja in, 1989). The provision of a wide range of contraceptive methods increases the opportunity for individual couples to obtain a method that suits their needs. A recent study of contraceptive method choice in the developing countries confirmed that prevalenc e is highest in countries where access to a wide range of methods is uniformly high (Ross et al., 2001).
SSRC Applications and Policy Working Paper A03/13Contraceptive choice is also a central element of quality of care in the provision of family planning services and an important dimension of women's reproductive rights (Bruce, 1990, Diaz et al., 1999. It is recommended that family planning programs offer a variety of safe, effective, acceptable and affordable contraceptive methods to help women prevent unwanted pregnancies and sexually transmitted diseases (STDs) and to help them achieve their childbearing goals. In addition, method-mix is a key determinant of the fertility impact of contraceptive practice; the use of more effective methods even by a smaller proportion of eligible