Assortative mating patterns have been little investigated in the developing world. Our paper examines the effect of changes in age at marriage and increasing schooling on the prevalence and conditions of unions in various contexts of the developing world. More specifically, we refer to the effects on assortative mating patterns with regard to education. We use census microdata for several countries available from the IPUMSInternational website. The selected countries are representative of various contexts of the developing world: Brazil 1991, 2000; China 1982, 1990; Iraq 1997; Kenya 1989, 1999; Mexico 1990, 2000; Philippines 1990, 2000; South Africa 1996, 2001; India 1993, 1999 (not presently integrated into IPUMS database). Results show some of the possibilities for comparative research of marriage patterns in time and space. We see that significant changes are underway as union formation, particularly for females, is postponed to the mid-twenties and beyond. At the same time, the proportions of never-marrying (or forming a union) increase. Educational attainment is an important factor in both these developments.