Environmental contaminants are ubiquitous and so are often key suspects in cases of lagging wildlife populations. How do we test hypotheses about cause-effect linkages between contaminants and wildlife health? We present three case studies in which different approaches were used to test hypotheses about effects of contaminants on wildlife. The cases involve the possible impacts of (1) polychlorinated biphenyl on Lake Superior bald eagles (Haliaeetus leucocephalus); (2) dioxin on osprey (Pandion halieatus); and (3) methyl mercury on common loons (Gavia immer). The different approaches were dictated by legal, logistic, and financial limitations, but the relative strengths of experimental and mechanistic approaches over correlative approaches is underscored. For all three species, the simple correlation between a single contaminant and performance was confounded by covariation with other types of contaminants and/or natural ecological factors such as food availability and predation