At the University of Calgary Faculty of Medicine in 1991, the authors sought to determine the effects of tutors' levels of content expertise on learning issues generated within problem-based learning (PBL) tutorials. For an integrative course taken prior to clinical clerkships, the 70 students in the class of 1992 divided into ten small groups, which were facilitated alternately by content experts and non-experts. The authors found that--across 35 simulated-patient case encounters (24 with non-experts and 11 with experts)--when the groups had tutors with expertise in the clinical cases studied, the groups generated approximately twice as many learning issues per case, and these issues were approximately three times more congruent with the case objectives. In addition, when the groups had expert tutors they spent approximately twice as much time per case in overcoming identified learning deficiencies. The authors conclude that it is important for tutors (1) to be well informed about cases and case objectives and (2) to be well versed in the PBL tutoring process.