One explanation for differences in treatment effectiveness for targeted symptoms is that more-effective treatments are more focused on patients’ problems than are less-effective treatments. This conjecture was examined meta-analytically. Comparisons of two treatments of adults with anxiety disorders were included. Effect sizes for targeted symptoms, nontargeted symptoms, and global outcomes (e.g., quality of life and well-being) as well as the relative focus on patients’ problems and researcher allegiance were coded. Metaregressions were conducted to predict effect sizes from (a) variables related to the focus on patients’ problems and (b) researcher allegiance. For symptom measures, the relative focus on patients’ problems predicted the relative effectiveness of the treatments, with the expectations created by explanation appearing more predictive than specific therapeutic actions focused on patients’ problems, although conclusions about relative importance were difficult to determine given collinearity of predictors. Researcher allegiance also predicted the effects of the comparisons. For global outcomes, both the focus on patients’ problems and researcher allegiance seemed to have smaller roles. A focus on patients’ problems appears to be important for the reductions of symptoms. Clinical trials comparing treatments need to balance the focus on patients’ problems and reduce researcher allegiance.