13 After approximately a half century of psy- chotherapy research, one of the most con- sistent findings is that the quality of the therapeutic alliance is the most robust pre- dictor of treatment success (Horvath & Symonds 1991; Martin, Garske & Davis, 2000). This has been evident across a wide range of treatment modalities. A related finding is that poor outcome cases show greater evidence of negative interpersonal process (i.e., hostile and complex interac- tions between therapists and patients) than good outcome cases (e.g., Henry et al., 1993). There is also a growing body of evi- dence demonstrating the relationship between resolving alliance ruptures and treatment outcome (e.g., Stiles et al., 2004; see Safran et al., 2002, for a review). Another relevant line of research has demonstrated that some therapists are con- sistently more helpful than others; differ- ences in therapist ability seem to be more important than therapeutic modality; and the more helpful therapists appear better able to facilitate the development of a ther- apeutic alliance (e.g., Luborsky et al., 1997). Process and Outcome Research The focus of our clinical research program is on clarifying the principles involved in resolving ruptures in the therapeutic alliance and developing training strategies to facilitate rupture resolution. A rupture in the alliance consists of a period of ten- sion, misunderstanding or lack of collabo- ration between patient and therapist. In this article, we will briefly describe some of our research findings as well as practice guidelines and training strategies that we have developed and tested for resolving ruptures in the alliance. The interested reader is referred to Safran and Muran (2000) for an extensive presentation of clin- ical theory, practice and training, and to Safran and Muran (1996), Safran et al. (in press), and Muran et al. (in press) for some examples of our research efforts. Over the last decade, we have investigated the processes through which alliance rup- tures can be resolved at various levels at analysis. At the level of psychotherapy process, we have devoted considerable effort to the study of alliance rupture reso- lution as a critical change event. This effort has included the development of assess- ment strategies to identify psychotherapy sessions in which such change events occur (e.g., patient- & therapist-rated post- session questionnaires), as well as those to assess the interactional sequences of vari- ous patient states and therapist interven- tions in the resolution process (e.g., observ- er-based measures of interpersonal behav- ior and emotional experience). Common elements of the resolution process identi- fied incude: 1) Therapist identifying and drawing attention to the rupture, 2) explor- ing patient feelings about the rupture that are being avoided, 3) exploring patient fears and expectations interfering with the expression of feelings about the rupture, and 4) patients expressing underlying wishes (Safran & Muran, 1996). At the level of overall outcome, we have conducted research evaluating the efficacy of a treatment approach that has been influenced by our process research as well as recent developments in relational psy- choanalysis. This approach is referred to as brief relational therapy (BRT). In one study FEATURE Brief relational therapy and the resolution of ruptures in the therapeutic alliance Jeremy D. Safran Ph.D., New School University J.C. Muran, Ph.D., Beth Israel Medical Center