The basic data needed for measurement of the risks of termination of the legal relationship of marriage by characteristics of the marital partners are not available at this time for the United States because the national divorce registration area includes less than half the States. Special studies based on selected census data or the records of marriages and divorces occurring in one State or community have provided much of the valuable but limited information at hand. Statistics for individual States are subject to substantial bias as a consequence of inter-State migration between time of marriage and time of divorce, but they must serve as a basic data source until national reporting has improved. A record linkage study was undertaken which tied marriages occurring in the State of Maryland in 1959 with divorces occurring in the State in the years 1959–66. Relative, not actual, divorce risks by race, age at marriage, and previous marital status were calculated for couples with at least one partner an in-State resident at the time of marriage. The dissolution rate was higher for whites than for nonwhites. Marriages contracted by persons at very youthful ages and by persons who had been married previously were found subject to greater than average risks of dissolution through divorce.