We evaluated hypotheses of the dynamics of predators (lynx) relative to prey (snowshoe hares) and predator abundance in the Yukon, Canada. The hypotheses were that predator (lynx) dynamics are influenced by prey density, or by both prey and predator densities. Annual lynx population growth rate (r), estimated from lynx counts, was positively related to previous hare density and negatively related to previous lynx density, as described by the best-fitting additive model (R² = 0.85). Annual lynx growth rate (r) estimated from lynx tracks was positively related to the ratio of hares per lynx in the best-fitting model (R² = 0.55). There was most support for the prey- and predator-dependent hypothesis of predator dynamics. Projected lynx tracks showed similar trends to observed abundance but lagged one year, emphasising the need for evaluation of projected predator trends.