This paper estimates labor supply elasticities of married men and women allowing for heterogeneity among couples (in educational attainments of husbands and wives) and explicitly modeling how household members interact and make labor supply decisions. We find that the labor supply decisions of husbands and wives are interdependent unless both spouses are highly educated (college or above). Couples with high education, the labor supply decisions of husband and wife are jointly determined only if they have pre-school age children. We also find that labor supply elasticities differ greatly between households. The participation own-wage elasticity is largest (0.77) for women with low education married to men with low education, and smallest (0.03) for women with high education married to men with low education. The participation own-wage elasticities for women with low education married to highly educated men and for women with high education married to highly educated men are similar and fall between these two extremes (about 0.30 for each). For all types of couples, participation non-labor family income elasticity is small. We also find that participation cross-wage elasticities for married women are relatively small (less than -0.05) if they are married to men with low education and larger (-0.37) if they are married to highly educated men. Allowing for heterogeneity across couples yields an overall participation wage elasticity of 0.56, a cross wage elasticity of -0.13 and an income elasticity of -0.006 for married women. The analysis in this paper provides a natural framework to study how changes in educational attainments and household structure affect aggregate labor supply elasticities.