Nonlethal sampling of bird blood and feathers are among the more common ways of estimating the risk of mercury exposure to songbird reproduction. The implicit assumption is that mercury concentrations in blood or feathers of individuals captured in a given area are correlated with mercury concentrations in eggs from the same area. Yet, this assumption is rarely tested. We evaluated mercury concentrations in blood, feathers, and eggs of marsh wrens in wetlands of Great Salt Lake, Utah, and, at two spatial scales, specifically tested the assumption that mercury concentrations in blood and feather samples from birds captured in a defined area were predictive of mercury concentrations in eggs collected in the same area. Mercury concentrations in blood were not correlated with mercury concentrations in eggs collected within the same wetland unit, and were poorly correlated with mercury concentrations in eggs collected at the smaller home range spatial scale of analysis. Moreover, mercury exposure risk, as estimated via tissue concentrations, differed among wetland units depending upon whether blood or egg mercury concentrations were sampled. Mercury concentrations in feathers also were uncorrelated with mercury concentrations in eggs, and were poorly correlated with mercury concentrations in blood. These results demonstrate the potential for contrasting management actions that may be implemented based solely on the specific avian tissue that is sampled, and highlight the importance of developing avian tissues as biomonitoring tools for assessing local risk of mercury exposure to bird reproduction.