This study was designed to identify the factors that influence an individual's length of employment in the field of nursing following graduation from a baccalaureate program. Subjects were classified as drop-outs or active, and compared with reference to personal, academic, and social variables. The data used for this study were compiled by the National Longitudinal Study of the High School Class of 1972. Subjects were measured at five separate intervals, beginning in 1972 and ending in 1979. Data collection instruments included the use of questionnaires and completion of a short test measuring verbal and nonverbal ability. Multiple correlation was used to identify those independent variables which were statistically linked. From this, a profile was developed and submitted to multiple regression analysis in order to define a prediction equation. A discriminant function analysis was performed to determine how well the variables were able to distinguish group membership. The results of this study indicate that academic variables are among the best predictors of length of employment in the field of nursing following graduation. Specifically, college grades are shown to be the best predictor in the early college years. As the student nears graduation, academic variables become less important and social variables become more important in the prediction of length of employment. Accurate prediction occurs in approximately 70% of the cases classified. These results would support the present practice of using grades as selection criteria in schools of nursing. While the social and personal variables did not have as much influence on length of employment as the academic variables, prediction efficiency was hindered by the number of variables under consideration and the small sample size. It is recommended that the study by replicated with a reduced number of variables and a larger sample.