Determining the means by which effective psychotherapy works is critical. A generally recommended strategy for identifying the potential causal variables is to conduct group-level statistical tests of treatment mediators. Herein the case is made for also assessing mediators of treatment outcome at the level of the individual participant. Single-participant assessment of mediators requires documenting, for each participant, that treatment was received, that change occurred on the mediator and relevant clinical outcome measures, and that the change on the mediator happened at an expected time in the treatment protocol and prior to substantive change on the dependent variable. Data from four depressed adolescents who demonstrated remission following a behavioral activation intervention illustrate the use of the approach in assessing whether changes in activation level or negative thinking mediated the changes in depression. For two participants, increased activation appeared to be a mediator, whereas decreased dysfunctional thinking never emerged as a plausible mediator. It is concluded that single-participant assessment of mediators of treatment outcome offers a useful additional tool for determining possible mechanisms of action in effective psychotherapy. (Contains 4 figures and 1 note.)