Three experiments addressed the role of response efficiency in the application of functional equivalence training. Functional equivalence training includes conducting a functional assessment of the problem behavior. Variables that predict and maintain the problem behavior are defined, and socially appropriate, functionally equivalent skills are identified and taught. The logic is that if the learner has a socially appropriate way to achieve the same function, he or she will be less likely to use problem behaviors. This study examined the role of response efficiency in functional equivalence training. Response efficiency was examined in terms of three variables: (a) physical effort, (b) schedule of reinforcement, and (c) the time delay between presentation of the discriminative stimulus and reinforcer delivery. Each of the three experiments involved a person who performed a set of problem behaviors and a functional assessment of the problem behaviors. A socially appropriate alternative response was taught, but this new response was less efficient than the problem behavior on one of the efficiency variables (effort, schedule, delay in time). The new behaviors did not compete successfully with the problem behaviors until a new, more efficient, alternative behavior was taught. These results are discussed in terms of our understanding of response covariation and the need in applied contexts to include response efficiency in any functional analysis assessment.