The U.S. Nuclear Regulatory Commission has developed a methodology to quantify the uncertainty in best estimate computer codes used to license commercial light water reactors. This methodology is equally applicable to other reactor designs with regard to providing a technical basis that supports the establishment and demonstration of compliance with safe operating margins. One of the cornerstones of the method is the identification and ranking of phenomena that are important to the postulated scenario. This paper references descriptions of the total methodology, describes the first three steps (i.e., through the identification and ranking of phenomena), and summarizes the results of the application of the methodology to a double-ended guillotine break loss of coolant accident in a production reactor.