A crucial step in assessing the scientific basis of a psychotherapeutic intervention is examining the individual components of the treatment to determine if they are additive or important to treatment outcomes. The construct of self-as-context (S-A-C), a central process in the acceptance and commitment therapy (ACT) approach, has not yet been studied in a component analysis. A previous dismantling trial, however, has shown this process has an additive effect as part of an ACT package (Williams, 2006). The current study is a preliminary trial of feasibility and efficacy to determine a) the practicality of assessing S-A-C in isolation in a laboratory setting, and b) the impact of manipulating S-A-C on theoretically related variables, including theorized mechanisms of change in various clinical approaches. 68 participants (55 female, 13 male) were randomly assigned to receive either a brief S-A-C intervention employing a common therapeutic metaphor (the chessboard metaphor), or the control condition, which involved discussing a mildly positive topic with the researcher. Results from the main analyses showed that there was no group-by-time interaction on measures to assess immediate impact on the construct, previously validated therapeutic mediation measures, or symptom measures. Several possible explanations for the failure to identify significant findings are discussed, including limitations of construct measurement. When analyses were repeated using only those participants whose scores were in the mild range or higher for stress, anxiety, or depression, time by condition interactions were significant for stress and approached significance for depression, with participants in the S-A-C group doing better than those in the control group, offering tentative support for the utility of this process among individuals with clinical difficulties. Implications for future studies are reported.