We assessed the usefulness of routine laboratory screening of preoperative patients. Computer-readable laboratory, demographic, and discharge diagnostic data were assembled for 2,000 patients undergoing elective surgery over a four-month period, and randomly selected samples of patients were studied. Several tests ordered by protocol and performed by the laboratory at the time of admission were examined in these samples, including complete blood cell count, differential cell count, prothrombin time, partial thromboplastin time, platelet count, six-factor automated multiple analysis, and glucose level. Sixty percent of these routinely ordered tests would not have been performed if testing had only been done for recognizable indications, and only 0.22% of these revealed abnormalities that might influence perioperative management. Chart review indicated that these few abnormalities were not acted on nor did they have adverse surgical or anesthetic consequences. In the absence of specific indications, routine preoperative laboratory tests contribute little to patient care and could reasonably be eliminated. (JAMA1985;253:3576-3581)