This series of experiments used rats to examine changes in behavioral control when stimuli with different associative histories are conditioned in compound. The initial experiments used blocking designs. Experiment 1 provided a within-subject demonstration of blocking, and Experiment 2 used the compound test procedure to show that, when a novel stimulus, X, is conditioned in compound with an already conditioned stimulus (CS), A, these audiovisual compound stimulus (AX)+ conditioning trials produce a greater increase in behavioral control for X than A. Experiment 3 showed that, when the blocked X is subject to further conditioning in compound with the blocking A (achieved by increasing the shock intensity on AX-shock trials), the compound trials again produce a greater increase in behavioral control for X than A. Finally, Experiment 4 showed that the unequal change in behavioral control for X and A was because of the difference in their training histories and not the test procedure. The overall pattern of results is consistent with the proposal (Rescorla, 2001) that associative change is regulated by the product of common and individual error terms rather than by common (Pearce & Hall, 1980; Rescorla & Wagner, 1972) or individual (Mackintosh, 1975) error terms. The pattern is also consistent with comparator theory (e.g., Miller & Matzel, 1988), which holds that the level of responding to a target is regulated by the strength of its comparator stimuli established in training.