We evaluated evidence of an effect of climate on the numerical response of a coyote (Canis latrans) population to their keystone prey, snowshoe hares (Lepus americanus), in a Canadian boreal forest. Six a priori hypotheses of the coyote numerical response were developed that postulated linear, nonlinear, additive, and interactive effects of prey and climate. Model selection procedures showed the North Atlantic Oscillation (NAO) had the strongest effect on the coyote numerical response via its interaction with snowshoe hare density, while other large-scale climate indices had very weak effects. For a given snowshoe hare density, a negative value of the NAO amplified the abundance of coyote and a positive NAO decreased coyote abundance. We hypothesize that the coyote numerical response is ultimately determined by the coyote functional response influenced by winter conditions determined by the NAO.