In an attempt to replicate an earlier study by H. G. Schmidt, H. P. A. Boshuizen, and P. P. M. Hobus (1988), 96 students and physicians at the University of Limburg (Netherlands) studied 4 clinical cases, produced a diagnosis, and recalled each case. Subjects were 24 second-year, 24 fourth-year, and 24 sixth-year medical students and 24 internists with at least 4 years of experience in internal medicine. In addition, processing time was manipulated. Diagnostic accuracy increased with level of expertise. Unexpectedly, the study failed to disclose an intermediate effect in the recall data. A positive, linear relationship between expertise level and case recall was found. Constraining processing time did not affect diagnostic accuracy, but equally affected the recall performance of the various expertise levels. The results are interpreted in terms of the inability of even expert physicians to process laboratory data in an automatic fashion. Two figures illustrate the report. (Contains 16 references.) (SLD)