Surveys on repressurization-induced backdrafting and spillage were conducted in threedifferent areas of the United States using a common protocol, primarily to assess thecorrespondence between short-term tests and one week of continuous monitoring per house.The short-term tests, under induced conditions, can only indicate whether there is a possibilitythat backdrafting or spillage might occur, whereas real-time monitoring under naturalconditions can give a true indication of backdrafting and spillage events.