Aptitude-treatment interaction (ATI) research is directed toward identifying client variables that predict differential treatment responsiveness. As such, insights derived from this research are of potential importance for clinicians faced with the dilemma of selecting the most effective treatment for their clients. Unfortunately, most of the ATI research to date has been characterized by the use of an inappropriate methodology. Therefore, some general methodological issues with regard to ATI research are outlined. The literature encompassing cognitive and/or behavioral treatments for anxiety, depression, pain, obesity, and tobacco dependence is then reviewed. This examination reveals little consistency in the emergence of individual differences relevant to treatment responsiveness. There are some indications that it may be possible to distinguish two (broadly denned) groups of individuals: those with active and those with less active coping styles. The former group appears to fare better in treatments that emphasize self-control rather than therapist control. In addition, the distinction between visually and verbally based coping strategies has been shown in some cases to predict response to visually or verbally based therapies. Overall, however, attempts to identify treatment-relevant client attributes have been disappointing, and suggestions for improving ATI research are forwarded.Over 2 decades ago, Kiesler (1966) attempted to dispel the myth of "patient uniformity," that is, the tacit assumption inherent in most psychotherapy research that patients with a given disorder constitute a homogeneous group. He argued that meaningful conclusions regarding outcome cannot be drawn until patient heterogeneity is recognized and incorporated into psychotherapy research designs. Thus, psychotherapy researchers must confront the question of "which therapist behaviors produce what changes in which kinds of patients" (Kiesler, 1971, p. 63). A minimal answer to this question requires a design that includes at least one group of subjects, subdivided on some relevant attribute, then randomly assigned to one of at least two forms of treatment. Aptitude-treatment interaction (ATI) studies have the capability of describing which treatment is best for which clients. However, "traditional" investigations continue to dominate the outcome literature. Studies that compare two or more different treatments are prevalent, despite the availability of appropriate methodological and statistical techniques for ATI research.Unfamiliarity with these methods has probably contributed to the slow growth of this type of research. Developments in ATI research also appear to be hampered by the lack of a unified effort to uncover the most promising client variables for Preparation of this article was supported in part by a Medical Research Council of Canada studentship to the first author and MRC Grant MA5028 to the second author. We would like to thank Lorrie Lefebvre for her help in preparing the manuscript and two anonymous reviewers for their ...