This paper argues that action research finds a rationale in the pragmatic position that knowledge is provisional and generated through a transaction between agent and environment. Action research finds a further methodological rationale in the pragmatic view that knowledge is generated within indeterminate situations, requires habits of reflection and analysis, and is arrived at through open agreement. However, pragmatic action research is also distinctive: it has a particular concern for consensus and, through the work of Dewey, a focus on the pedagogical implications of problem solving. This paper discusses the value of the label "pragmatic" and the strengths and weaknesses of the pragmatic approach.