Abstract: The therapeutic alliance has been accepted as a central component of the psychotherapeutic process. Numerous studies have been conducted on theories, measurement issues, and correlates of the alliance, however, researchers only recently have started to acknowledge the possibility of improving the initial development of the alliance. The present studies investigated three facets of the alliance, using it to predict outcome, examining its pretreatment predictors, and using inferential feedback in a pilot study to enhance its development. Results showed that the alliance significantly predicted outcome as measured by level of client distress, experience of overall symptoms, and unusual thinking. Client intake relationship impairment and racial match significantly predicted the quality of the alliance. Finally, there was no statistical support provided for the efficacy of the inferential feedback intervention conducted in the pilot study. Suggestions are made for improvements to the feedback intervention. In addition, directions for future research on the therapeutic alliance are discussed.