In the first set of experiments, aversion of rats to a salty taste induced by ingestion of the LiCl solution in one context was extinguished by repeated exposure of animals to an NaCl solution of constant concentration or by gradually fading an NaCl-sucrose mixture in another context. In the end, the rats were tested with the NaCl solution in the original context.Their performance was compared with that of control rats that were tested in the same, unchanged context. As explained below, the results of the first experiment (Experiment 1A) were unexpected, and thus, the experiment was repeated twice with some minor modifications (Experiments 1B and 1C). Contrary to our expectation, all of these experiments (1A, 1B, and1C) yielded numerically larger renewal effects after graded, rather than constant, extinction treatment. Hence, in Experi-ments 2 and 3, we compared the extent of spontaneous recovery and reacquisition between graded and constant extinction groups, respectively. In general, responding extinguished by the graded procedure was more susceptible to relapse than conventionally extinguished responding.