Influences on students' reported academic development over a 4-year period were assessed, focusing on college influences on student growth and changes each year. Tinto's model of college student attrition was found to account for about 23% of the variance in students' reports of their academic skill development over 4 years of college. For each year's follow-up instrument, students indicated two things: the estimated number of times during the year they had met with a faculty member outside the classroom for at least 15 minutes; and the average number of hours per week they had spent in organized extracurricular activities. Students also completed questionnaires that measured social and academic integration, as well as classroom and social involvement. LISREL analysis indicated that students' academic integration in each of the 4 years had a direct and indirect effect on reported academic skill development in that year and in succeeding years. Social integration was more influential in students' reported academic growth in the junior and senior years. The nature and strength of these two influences varied over time, however, with the relative importance of academic and social integration reversing over the period. (SW)