Computerized information systems are viewed as innovations in the health care delivery system and are used by nurses to support the management of clinical and administrative information. The development, implementation and use of computerized information systems are viewed as predictable activities in the future health care environment. Adoption of these innovations is critical to the achievement of improved information management. Nurses are challenged to direct these activities to achieve efficiency, effectiveness and productivity in the delivery of health care services. The adoption of information systems is envisioned as a vehicle to enhance and improve the quality of information and ultimately the quality of nursing practice. This study explored factors that influence the adoption of a computerized information system as an innovation after it is first introduced to members of a social system. Individual, technological and organizational characteristics were investigated as predictors of adoption using multiple regression analyses, and the effect of organizational position on adoption was analyzed. Findings evidenced that over one half of the variance in adoption was explained by all three sets of predictors; close to one half of the variance was explained by technological variables alone; and one third of the variance was explained by organizational variables alone. Factors related to an adopter's personality, socio-economic status and communication behavior were found not to influence nurse adoption. In addition, using analysis of covariance, a significantly lower level of adoption was found for managers compared to non-managers when either need, supervisor values or peer values was controlled. The results suggest that focus on the perceived advantages of an innovation, the perceived need, values held by peers regarding the innovation and use of communication mechanisms facilitate nurse adoption of a computerized information system as an innovation. A revised theoretical model for the study of innovation adoption was proposed based on the data analysis. The model asserts that a relationship exists among the categories of predictor variables and negates the hypothesized direct influence of adopter characteristics on adoption behavior. Recommendations for further research, replication and model testing are proposed to expand the body of knowledge in this field. Implications for nursing include the need to focus strategies for planning, development and implementation of computerized information systems on maximizing the significant predictors. Evaluations of the information management component of current nursing roles is also suggested.