Three successive groups of education students enrolled in a fifth year M.Ed. program were tested in a longitudinal study for differences between their attitudes after completing the academic portion of the program (first semester) and after completing internship (second semester), which involved full teaching responsibility along with biweekly supervision by the college and a weekly seminar. The Minnesota Teacher Attitude Inventory (MTAI) was administered in September, January (when roles changed from student to intern), and May. Analysis of MTAI scores for 52 subjects revealed a significant gain in scores (i.e., in tolerance and child-centeredness) after completion of the academic program and a significant decrease in scores after internship. Tentative support was found for the hypothesis that there is more variability in the direction and amount of change during internship than during the academic program. When MTAI scores were correlated with intern ratings made by the program director of the amount of difficulty and frustration each intern experienced, the latter were found to be significantly associated with a decrease in MTAI scores. Any hypothesized relationship between internship in an inner-city school and a decrease in score approached significance. Variables and hypotheses suggested by this investigation need to be explored. (A summary of the preceding four-year pilot study is included.) (LP)