The majority of psychotherapy practitioners use therapist self-disclosure (TSD; Lane, Farber, & Geller, 2001; Henretty & Levitt, 2010), clients say it is helpful (Hanson, 2005) Hill, Helms, Tichenor, Spiegel, O’Grady, & Perry, E., 1988), and a growing body of research and theory suggests that avoiding TSD in all circumstances may have harmful effects on both the client and the therapy (Barnett, 2011). Thus, continued research is called for to provide clinicians with recommendations for how to use the intervention therapeutically, as well as how to avoid using it in ways that might be harmful. However, little is known about how master therapists make decisions about TSD, and researchers have found that studying therapists’ use of disclosure, in general, is of limited use (Gallucci, 2002). Accordingly, we interviewed 13 master therapists about their general attitudes about TSDs, examples of actual successful TSDs, examples of actual unsuccessful TSDs, and instances during which they felt an urge to disclose but chose not to do so. We analyzed the transcripts using consensual qualitative research. In terms of general attitudes, therapists believed that some types of TSD can be helpful in some situations if used sparingly, but had many cautions about using TSDs. In successful TSDs, there were no typical antecedents; therapists typically intended to provide support, facilitate exploration and insight, and build and maintain the therapeutic relationship; the content was typically about similarities between the therapist and client and relevant to the client’s issues; and the consequences were typically positive. In unsuccessful TSDs, the typical antecedents were countertransference reactions; the typical intentions were to provide support; therapists typically misjudged perceived similarities; and the consequences were negative. In instances when therapists felt urged to disclose but did not, the typical antecedent was countertransference; and the content of what was not disclosed typically seemed relevant to the client’s issues. Implications for practice, training, and research are discussed.