The main features of courses taught according to the Personalized System of Instruction (PSI) as described by Keller are reviewed, along with the use of repeated testing and undergraduate tutors. The PSI approach is also evaluated in terms of final exam performance, study behavior, and student ratings, based on a study conducted at York University, Ontario, Canada. The PSI courses involve: infrequent lectures and demonstration; an emphasis on written materials divided into units or modules; a clear statement of objectives preceding each unit; mastery of objectives as evidenced in performance on criteria-referenced unit tests that are repeated if necessary; and undergraduate proctors/tutors who administer quizzes and other services. The effect of three methods of evaluating students' term work on their course performance was assessed: (1) repeated testing (i.e., students were given dates for a number of evaluation exercises that they were allowed to repeat 1 week later); (2) no repetition of exercises was permitted but students were given dates for the evaluation exercises and were told that each was an optional exercise; and (3) students were allowed complete freedom to determine the nature and frequency of the evaluation exercises that they would submit. Results were not definitive, but additional research was conducted on the use of undergraduate tutors. Perspectives on the ways in which a course can be individualized and which allows variations in: the content studied, when students will study the material, the behavior of the instructor, and the method of evaluation. Ways that undergraduate assistants can help promote these variations are discussed, and self-pacing and mastery are briefly addressed. (SW)