A causal model to explain student attrition was tested at a major midwestern land-grant university with a sample of 1,513 full-time, unmarried freshmen who were 21 years old or younger. The causal model was reduced from 23 to 10 variables: an intent variable, three attitudinal variables, and two each of organizational, personal, and environmental variables. Background variables were excluded. The sample was divided into four groups based on the student's sex and level of self-confidence, and multiple regression and path analysis were used to analyze the data. Overall ranking of the independent variables in explaining dropouts, based on effects coefficients, in descending order of importance, was as follows: intent to leave, grades, opportunity to transfer, practical value, certainty of choice, loyalty, family approval, courses, student goals, and major and job certainty. The correlation coefficients ranged from .42 to .50. For each of the four path analyses (high/low confidence women and high/low confidence men), intent to leave had the largest direct influence on dropping out. Also, for each group, the three attitudinal variables (loyalty, certainty, and practical value) had significant negative relationships with intent (with the exception of loyalty for low confidence men). However, the attitudinal variables were not well explained themselves due to relatively high intercorrelations. Recommendations based on the findings are as follows: develop the motivation and learning skills of students so that their grades can rise; demonstrate to students how any major they choose can be of practical value (important for future employment); create a desirable image of the school and identify reasons to be loyal to it; offer courses the students think that they want to take; and develop the student's educational goals. (SW)