Psychotherapy process includes insession process and intersession process, occurring in and outside the counseling room respectively, and the latter has been neglected for a long time in psychotherapy research history. Intersession process refers to the thoughts, memories, and feelings about therapists or therapy sessions that clients experience during the intervals between sessions. Intersession process is a joint product of actual in-session therapeutic interaction and the client’s prior personality, in particular, the client’s conscious and unconscious transference potentials. It connects the therapeutic process and client’s actual life, and has repaired functions for clients in therapy. The present study aimed to identify the effects of intersession processes on treatment outcome over successive sessions in short-term therapy. 12 male and 24 female student visitors of a college counseling center, with a mean age of 22.47(SD=2.60) and without personality disorder or other psychiatric symptoms, participated in the study. Intersession process was measured by Chinese version Intersession Experience Questionnaire (IEQ), which consisted of eight factors of four dimensions. Treatment Outcome was measured by Outcome Questionnaire-45 (OQ-45) and higher scores meant more severe symptoms. Participants were asked to complete IEQ and OQ before each session from session 2 to 9. Finally 151 sessions/36 cases were collected, and the mean measures of each case was 4.19 (SD=1.26). A two-level hierarchical linear model was built to investigate the effects of time and time-depended predictor IEQ on OQ. Results indicated that, generally, in the short-term psychotherapy the symptoms of clients significantly declined in linear tendency, while OQ decreased by 1.27 point after each session. The proportion of total variance between the cases of the regression equation was 79.76%, which meant the necessity to use multilevel analysis. The initial symptoms and the change rates of OQ over time differed significantly among different cases. The different change rates could be well predicted by the positive and negative factors of emotion dimensions of IEQ, with contributions of 40.79% to total rate variance. When IEQ positive or negative emotion factors increased by 1 point, the OQ scores decreased by 7.398 or increased by 6.433, respectively. However, other three dimensions of IEQ had no significant effects on outcome. The results suggested that the intersession processes could well predict the different change rates of symptoms among individuals in short-term therapy, and intersession emotions may reflects the client’s transference potentials to therapist, which can be acknowledged and used to facilitate treatment.