The authors challenge the assumption that the passive lecture-based curriculum that dominates most preclinical medical education optimizes learning for all students. For this reason, an experiment was conducted in which medical students were given the option of taking some courses as small groups in a Socratic format rather than in lectures; 67 percent of the students applied for the small-group option. Personality factors, measured by the Myers-Briggs Type Indicator, were used to determine whether there was any correlation between learning preferences and personality types. Most of the students preferred having more opportunities for small-group instruction. Most students and faculty members found the experience profitable; a significantly greater proportion of students of the intuitive-feeling personality type reacted positively to the experience than did other personality types. The small-group classes required more faculty time than did lectures, and this often prevented faculty support of the Socratic approach.