Social power involves the potential to influence others and is important in explaining organisational outcomes. Based on the assumptions posited in Raven’s (1992, 1993) power interaction model, we examined a set of potential antecedents and correlates of power sources. Participating in the study were 232 nurses and 32 supervisors from two municipal hospitals. Nurses and supervisors were asked to complete a battery of questionnaires including demographic items, the Interpersonal Power Inventory, job satisfaction, and organisational commitment. Convergent and discriminate validity of the power inventory yielded satisfactory results. Principal component analyses of the power inventory revealed a two-factor solution (harsh/soft sources) which was then used for comparison with predictors and outcome data. Among the major findings were the fact that compliance to power sources appears to be contingent on objective and subjective indices of professional distance between supervisors and subordinates, supervisor seniority in the department, and the type of promotion (from without or within department). Job satisfaction was found to be positively related to compliance with soft sources and negatively with harsh sources whereas commitment was positively associated with both.