A new approach to the prediction of electromagnetic compatibility between possible culprit and victim cases is introduced. It is shown that this approach, based on statistical descriptions of case emission and case susceptibility, leads directly to the determination of the probability of interference between randomly oriented equipments in operational situations. Results of tests made using a technique to statistically describe case emissions indicate that greater measurement accuracy and repeatability can be obtained with this technique than is inherent in the case emission measurement methods currently in use and recommended by most specifications. Measurements made on culprit sources in a typical laboratory work area using the statistical technique were found to routinely yield results within + 1.5 dB of corresponding values measured in a free space or anechoic chamber environment. When the technique was used with a hooded probe antenna, equally precise results were obtained in a shielded enclosure. Used in conjunction with appropriate statistical descriptions of case susceptibility, case emission data such as those accumulated with the statistical technique discussed could be used to predict the probability of electromagnetic compatibility in actual equipment lashups.