Research in psychotherapy and clinical psychology is following increasingly the medical model, trying to isolate “specific effects.” This posits that only such “specific effects” are important, while nonspecific or “placebo-effects” are neglected. However, the size of these nonspecific effects can vary between treatments. This situation leads to the efficacy paradox: Treatments that have difficulties showing statistical superiority over placebo controls can nevertheless be more effective than other types of treatments that have shown clear superiority over placebo controls. This paradox is introduced here, and the conditions are analyzed. The efficacy paradox is likely important for any type of research that tries to distinguish between “specific” and “nonspecific” effects and thus questions this distinction.