One experiment each was conducted with pigeons and rats to assess the effects of changes in reinforcer magnitude on resurgence. Each experiment involved three phases. In the Training phase, key pecking (Experiment 1) or lever pressing (Experiment 2) on two concurrently available operanda was reinforced according to variable‐interval schedules. In the Alternative Reinforcement phase, responding to one operandum was extinguished while that to the other was reinforced with greater duration of food access (Experiment 1), greater number of pellets (Experiment 2a), or a similar number of pellets (Experiment 2b) than occurred in the Training phase. In the Resurgence Test phase, the reinforcer magnitude associated with the Alternative response was either reduced (Experiments 1 & 2a) or increased (Experiment 2b) relative to the preceding condition. Resurgence generally occurred when the reinforcer magnitude maintaining the Alternative response was reduced, but not when it was increased relative to the preceding condition. The results further support the suggestion that resurgence results from an overall “worsening” of reinforcement conditions, but not simply from a change in conditions.