The construct of "role conflict" is given considerable space in contemporary sociological thinking, yet a review of the literature in the area would seem to suggest two things. Firstly, it is apparent that authors are using the construct in different ways, and hence giving the same term different meanings. Secondly, it is possible that greater explanatory and predictive power might be obtained from "conflict" measures beyond those traditionally used. A conceptual framework is offered in which the relationships between different authors' measures can be seen, and from which additional approaches to the construct might be made. Use is then made of this framework to generate 162 different operationalizations of the construct using data from an unrelated study on teacher role. These measures are then tested for their relative utility against three criterion measures. Further, they are mathematically manipulated in order to obtain a compound operationalization with considerable explanatory power of the variance of one of the criterion measures. The study is essentially a methodological argument, so that no substantive conclusions about role conflict in teachers are made.