A retrospective study was done with 325 patients who had preadmission testing prior to ambulatory surgery. At least one laboratory abnormality was noted in 84% of the patients. The serial multiple analysis (SMA)-7 was abnormal 63% of the time. Abnormalities were seen in 54% of the SMA-12 panels and 38% of the urinalyses performed. Twenty-four percent of the patients treated had an abnormal electrocardiogram (ECG). An abnormal chest roentgenogram was found in 19% of the patients. Only three (1%) patients potentially benefited from preadmission testing. Ninety-six percent of the abnormal laboratory results were ignored by the attending physicians. Therefore, we conclude that preadmission testing should be done on a selective basis. Patients older than 50 years of age should have an ECG. A hematocrit should be obtained only if major blood loss is anticipated. All other tests should be ordered based on the history and physical examination.