Water quality criteria (WQC) can be derived in several ways. The usual techniques involve hazard and risk assessment procedures. For non-persistent, non-biomagnified compounds and elements, WQC are experimentally derived from their acute and chronic toxicity to aquatic organisms. For those persistent chlorinated hydrocarbons (PCHs) that are bioaccumulated and biomagnified, these traditional techniques have not been effective, partly because effects higher in the food web were not considered. Polychlorinated biphenyls (PCBs) are the bioaccumulative synthetic chemicals of primary toxicological significance to the Great Lakes biota which have caused widespread injury to wildlife. In the Laurentian Great Lakes, the primary emphasis of hazard assessments has been on the potential for adverse effects in humans who eat fish. The primary regulatory endpoint of traditional hazard and risk assessments underlying current WQC are the probabilities of additional cancers occurring in the human population. The analysis presented here indicates that this is not adequate to restore sensitive wildlife species that are highly exposed to PCBs, especially those that have suffered serious population declines. Because WQC are legal instruments, the methods of deriving WQC have large implications for remediation, litigation, and damage assessments. Here WQC are derived for six species based on the responses of wildlife in the field or produced by feeding fish to surrogate species, rather than projecting a potential of increased cancer rates in humans. If the most sensitive wildlife species are restored and protected for very sensitive reproductive endpoints, then all components of the ecosystem, including human health, should be more adequately protected. The management of Great Lakes wildlife requires an understanding of the injury and causal relationships to persistent toxic substances.