This study makes use of the National Family Health Survey of 1998-99 to investigate whether differences in women's autonomy can explain much of the relationship between education and contraceptive use among married Indian women with at least one child. The analyses show that a woman's education does not influence her contraceptive use through a strengthening of her position in relation to that of men, but that the inclusion of a simple indicator of her general knowledge reduces education effects appreciably. Further, the average educational level of other women in the census-enumeration area has an effect on a woman's contraceptive use above and beyond that of her own education. This effect cannot be explained by the specific indicators of autonomy, but can to some extent be explained by the son preference of the community. The latter is a more general autonomy indicator that may also pick up other contextual factors.