Research has shown that women are underrepresented in positions of educational authority. This paper presents findings of a study that asked the following question: What is it about the regularities in discourse and practice in relationship to power in a particular community that would allow a woman to be selected for the superintendency, when about 96 percent of the time a man is selected? It is hypothesized that the definition of power is gender-specific (i.e., women define power as collaboration or "power to," whereas men define it as domination, or "power over"). Using Stewart Clegg's (1988) conceptualization of power relationships as "circuits of power," this critical ethnography was conducted in a larger metropolitan area headed by a female superintendent. Data were obtained from a total of about 100 nonstandardized interviews with those in male circuits of power, those in female circuits of power, and those related to the superintendent either directly or indirectly. Other data sources included document analysis and nonparticipant and participant observation. Findings indicate that: (1) power is defined differently by women than men; (2) when women operate according to the female concept of power, their chances to acquire positions of power increase dramatically; (3) women are most likely to be empowered in those communities that have pluralistic or diffused power structures; and (4) women who attain positions of power are most successful when they adopt female approaches to power that stress collaboration, inclusion, and consensus-building models. Contains 68 references. (LMI)