The authors investigated the hypothesis that the therapeutic alliance mediates the relationship between pretherapy expectancy of improvement and psychotherapy outcome. Data were drawn from a comparative trial of 2 forms of short-term, time-limited individual psychotherapy (W. E. Piper, A. S. Joyce, M. McCallum, & H. F. Azim, 1998). Measures of expectancy and outcome were based on an individualized assessment of target objectives; outcome was considered from 3 perspectives (patient, independent assessor, therapist). Using the R. M. Baron and D. A. Kenny (1986) procedure, the authors found evidence in support of the hypothesized mediation effect. The effect was evident when the alliance was rated from the perspective of either patient or therapist, and it accounted for one third of the direct impact of expectancy on outcome. Clinical implications and limitations of the study are discussed. [ABSTRACT FROM AUTHOR]