Proprietary colleges are uniquely different from two or four year colleges due to the emphasis on the student establishing a definite career path prior to enrollment. Because of this career track emphasis, Bandura's (1977) postulation that self-efficacy is a significant variable influencing task completion may offer insight into the challenge of student retention at a proprietary college. The study's purpose was to determine if career self-efficacy, demographic factors, and academic preparedness measures in first quarter students could predict student persistence, class attendance, and academic performance. The statistical technique of multinomial logistic regression was applied to data files of 725 first quarter students who attended The Art Institute of Dallas from Summer 1996 through Winter 1997. The predictor variables included a measure of career self-efficacy, ASSET scores (American College Testing Program, 1994), ethnicity, age, gender, full-time/part-time attendance, high school grade point average, parents' educational level, socioeconomic status, and developmental course placement. Criterion variables were completion, class attendance, and cumulative grade point average.