Discourse (DA) and conversation (CA) analysis, two qualitative research methods, have been recently suggested as potentially promising for the study of family therapy due to common epistemological adherences and their potential for an in situ study of therapeutic dialog. However, to date, there is no systematic methodological review of the few existing DA and CA studies of family therapy. This study aims at addressing this lack by critically reviewing published DA and CA studies of family therapy on methodological grounds. Twenty-eight articles in total are reviewed in relation to certain methodological axes identified in the relevant literature. These include choice of method, framing of research question(s), data/sampling, type of analysis, epistemological perspective, content/type of knowledge claims, and attendance to criteria for good quality practice. It is argued that the reviewed studies show "glimpses" of the methods' potential for family therapy research despite the identification of certain "shortcomings" regarding their methodological rigor. These include unclearly framed research questions and the predominance of case study designs. They also include inconsistencies between choice of method, stated or unstated epistemological orientations and knowledge claims, and limited attendance to criteria for good quality practice. In conclusion, it is argued that DA and CA can add to the existing quantitative and qualitative methods for family therapy research. They can both offer unique ways for a detailed study of the actual therapeutic dialog, provided that future attempts strive for a methodologically rigorous practice and against their uncritical deployment.