Methods for evaluating a single diagnostic test with reference to the disease prevalence in a given population include sensitivity, specificity, positive predictive value, negative predictive value, and accuracy. Patient disease status and, ultimately, the treatment course is determined by the outcome of these diagnostics. The advantage of ordering a single diagnostic test, or series of diagnostic tests is a concern of physicians. How much information is gained from using the results of two diagnostic tests, each designed to detect the same disease? Combining tests may be the optimal methodology for determining the disease status of the patient. We propose a systematic strategy for optimizing (minimizing alpha and/or beta errors) the combination of two diagnostic tests. This strategy is then illustrated by the use of data from Doppler ultrasound and ocular pneumoplethysmography in detecting carotid artery disease.