This stLtdy hypothesized that family role (marital and porentnl status) would moderate the effect of disaster exposure on the meiital health of' victims. The study included St. Louis residents exposed to jloods and dioxin, as well as Puerto Rican respondents erposed to floods and mudslides. In St. LOU~S, worst outcomes were found for single and married parents exposed to disaster, substantially wceeding the symptomatology of all unexposed respondents except non-victim single parents. In Arerto Rico, victims witliout families had higher levels of alcohol abuse symptoms than did any other subgroup. Perceived emotional support was found to be an important moderator of disaster's effect on psychiatric distress in this site, generally overriding the effect of family role. Single parents in both sites who were exposed to disaster had substantially reduced levels of emotional support available to them, us compared to unexposed single parents, suggesting that single parents are at particularly high risk for losing access to emotional support following a disaster. This study suggests that both single and married parents constitute important high-risk victim groups. The findings also suggest that those perceiving they lack adequate emotional support, regardless of family role, may be in special need of services.