Most studies of caregivers are limited to nonrepresentative samples, samples only of women, or samples of caregivers nominated by care recipients. This makes it difficult to place caregiving in a population context, to analyze differences between caregivers and noncaregivers, and to explore possible gender by caregiving interaction effects. This study used nationally representative data from the National Survey of Families and Household 1987-88 to: 1) describe the prevalence of caregiving among midlife men and women (ages 35-64, N=5,643); 2) examine differences between caregivers and noncaregivers in health, psychological well-being, social participation, and marital quality; and 3) examine evidence for differences between men and women in the effects of caregiving on these outcomes. Almost one in five midlife adults were found to have recently been involved in caregiving either in or out of their residence. The ratio of female to male caregivers in this study was about four to three -- indicating a much higher prevalence of male involvement than most nonrepresentative sample studies indicate. Few health and well-being effects of caregiving were found to differ by gender; there were no differences between men and women in caregiving effects on depression. The biggest differences between caregiving men and women were evident in marital quality -- more disagreement, less sex, less activity, and a higher predicted likelihood of having their marriage end reported by women.