Fish species at risk (n = 29) and not at risk (n = 88) of extinction in the Great Lakes –St. Lawrence biozone were compared to determine if they could be distinguished by their life-history characteristics. A matrix of 51 variables representing 2 phylogenetic, 18 life-history, and 31 ecological variables was compiled from literature sources. Nonparametric analysis shows that 14 variables distinguish between species at risk and not at risk. Logistic regression indicates that seven factors are significant in defining the two groups: age at maturation, feeding habitat, fish feeding regime, feeding substrate, water flow over feeding area, breeding habitat, and breeding substrate. The resulting model is concordant at 97.1% (p = 0.0001, −2 log likelihood = 38.9). The model states that any fish species maturing at 18 years or more is at risk of extinction in the biozone. When applied to a set of fish species from the American Midwest using a decision level of p = 49%, the model had a reclassification success of 77.8%. The modeling technique could be useful to evaluate species at risk and to complement decision making with regard to vulnerability, especially in the case of wide-ranging species such as Atlantic sturgeon (Acipenser oxyrhynchus) and American eel (Anguilla rostrata).