Medical students and practitioners usually learn how to use problem-oriented medical records (POMR) by haphazard means. In attempting to put system into the teaching of POMR, the authors devised teaching case material and two instructional formats: self-instruction and workshop. The results of a controlled study with second-year medical students to determine the relative effectiveness of instructional formats and case materials are presented. At the end of instruction POMR performance was measured by assessing the student's ability to convert a case to POMR format. The results suggest that all student groups attained an acceptable performance level no matter which format they experienced.