An analysis of data from the 1973 and 1976 National Surveys of Family Growth indicates that although marital disruption affects contraceptive practice, the patterns of contraceptive behavior among separated women who remain sexually active are very similar to those of women who remain married. Women about to separate rely less on barrier methods than do continuously married women, and this pattern persists after separation. It is likely that coitus-related methods are less frequently used because they require cooperation between the spouses. In addition, coitus-related methods are probably used less after separation because these methods are less effective, and unintended pregnancies are less desirable in the intermarital or postseparation period. As might be expected, both before and after separation, women who experience marital dissolution are less likely than the continuously married to be sexually active. The difference is greater after separation, but the major finding is the similarity of separated and continuously married women with respect to both coitus and contraceptive protection.