Data from the National Survey of Families and Households for 1987-1988 are used to explore methodological and substantive issues concerning marital dissolution in the United States. "The analysis finds that marital disruptions are seriously underreported by males, making the analysis of male marital histories problematic. Also, the potential impact of reconciliations on the estimates of recent marital disruption based on separation is explored; no upward bias is likely to result from the inclusion of separations that may subsequently reconcile. The impact of a wide variety of factors on the risk of marital disruption is examined using proportional hazard techniques. Among them are included parental background factors, respondent's characteristics at the time of marriage, differences in spouses' characteristics, and joint activity statuses of marital partners in the first year of marriage. The risk of marital disruption is highest among women with young age at marriage, low education, a cohabitation history, and those whose spouse has been married previously. Parental family disruption affects marital stability primarily through age at marriage and cohabitation. Religious and educational heterogamy and male unemployment reduce marital stability."