Context  The Liaison Committee on Medical Education (LCME) requires there to be: ‘…comparable educational experiences and equivalent methods of evaluation across all alternative instructional sites within a given discipline’. It is an LCME accreditation requirement that students encounter similar numbers of patients with similar diagnoses. However, previous empirical studies have not shown a correlation between the numbers of patients seen by students and performance on multiple-choice examinations.

Objective  This study examined whether student exposure to patients with specific diagnoses predicts performance on multiple-choice examination questions pertaining to those diagnoses.

Methods  The Department of Pediatrics at the University of Nebraska Medical Center has collected patient logbooks from clerks since 1994. These contain information on patient demographics and students' roles in patient care. During week 7 of an 8-week course, students took an examination intended to help them prepare for their final examination. Logbooks and pre-examination questions were coded using standard ICD-9 codes. Data were analysed using Minitab statistical software to determine dependence between patient encounters and test scores. Subjects comprised a convenience sample of students who completed the clerkship during 1997−2000.

Results  Our analysis indicates that performance on a multiple-choice examination is independent of the number of patients seen.

Conclusions  Our data suggest knowledge-based examination performance cannot be predicted by the volume of patients seen. Therefore, too much emphasis on examination performance in clinical courses should be carefully weighed against clinical performance to determine the successful completion of clerkships.