This paper explores possible uses of teacher efficacy research by democratic teacher educators. The many possible meanings of teacher efficacy levels make it problematic for teacher educators to interpret and use teacher efficacy. Beneficial teacher efficacy doubts and problematic types of teacher efficacy confidence are identified, including conceptions of agency that may conflict with democratic teaching. Thus, efforts to enhance teacher efficacy may have many negative side effects. The research support for teacher efficacy is critiqued. To make future research more useful to democratic teacher educators, a new definition of teachers' efficacy beliefs is proposed, and research directions are outlined.