Two versions of a counselor self-efficacy (CSE) measure were administered to 110 prepracticum counselors: a general version, assessing perceived capability to perform basic helping skills and manage the session process with clients generally; and a client-specific version, tapping capability to perform the same behaviors with a specific, current client. Client-specific CSE was found to (a) relate moderately to strongly with general CSE over the course of four counseling sessions, (b) increase significantly over sessions, and (c) account for unique variance in counselors' evaluations of the quality of their sessions. Although it was not a useful direct predictor of clients' session ratings, higher client-specific CSE was associated with greater congruence between counselors' and clients' perceptions of session quality. Implications for further research and training are considered.