This study examines the social networks of older widowers in small town and rural settings in Iowa by comparing men in three marital statuses--married, widowed, and formerly-widowed-but-now-remarried. Findings indicate that men in these categories differed in ties to some social networks, but widowers relative to married and remarried peers were not socially isolated;Noting that married, widowed, and remarried men have divergent social network ties does not mean that differences are explained by marital status. Exchange and network approaches are used to develop a model for testing relative contributions of marital status and a series of intervening variables for predicting network involvement and consequences for well-being. Intervening variables consist of three sets of variables--characteristics of the networks, individual ties to networks, and personal resources. Multiple regressions reveal that marital status explains little variance in child interaction, friend interaction, or life satisfaction once intervening variables are considered. Network characteristics are crucial for explaining interaction; network ties, and especially personal resources, are important predictors of life satisfaction;Findings suggest that future research is needed. The exchange-network model should be tested using widows and urban widowers to ascertain whether marital status is of minor importance in explaining involvement and well-being among other populations. A second need is for longitudinal research with widowers to determine whether continuity or compensation occurs in network ties to prevent social isolation with widowerhood. Third, multiple confidant ties, net worth, and dependency--three theoretically important variables rarely included in research--require additional examination before generalizations can be reached regarding their importance as predictors of network involvement and consequences for well-being.