Remarriage is one of the most important determinants of physical and economic well-being among the widowed. The goal of this study is to estimate how hazard rates for remarriage vary among widows and widowers on the basis of both observable and unobservable characteristics. The remarriage estimates rely on nationally representative samples of widows and widowers from the Panel Study of Income Dynamics. Continuous-time hazard rate regressions indicate strong age and duration dependence effects for middle-aged widows and widowers and for older widowers. Among middle-aged widows, blacks and those with dependent children in the home have lower rates of remarriage. For middle-aged widowers, living in urbanized areas limits the prospects of remarriage. For older widowers, education and, to some extent, economic status appear to have positive effects on the remarriage rates. Overall, age and time since widowhood have the strongest and most consistent effects on remarriage rates for different widowed groups.