IntroductionTo construct appropriate boundary conditions for wavebased room acoustics simulations, the authors have proposed a method for measuring sound absorption of materials using the ensemble averaging technique [1,2]. We have also shown several applications of the method (EA method, hereafter) both conditions in the laboratory and in situ [3,4].Based on several example calculations conducted for a music hall with volume of 11,000 m 3 , Vorländer pointed out that the uncertainty of the absorption coefficient must remain less than 0.04 to keep the uncertainty of calculated reverberation time below its just noticeable difference (JND) [5]. There followed the statement that such a small uncertainty cannot be obtained from the reverberation room method following ISO 354. Although Vorländer made no statement related to the tube method, the mounting problem [6] of the method has not been solved yet. It is likely to affect the resulting uncertainty as expressed by Stanley [7] that ''an impedance tube system to measure the acoustic absorption is not an extremely precise and repeatable process because of unavoidable variations of specimen cutting and cell fit.'' Herein, as an alternative method, the uncertainty of EA method is discussed. First, theoretical basis of EA method is described briefly, proposing a mathematical-physical model of ensemble averaging to clarify the definition of measured absorption with respect to sound incident events. The knowhow is described to measure the absorptions of reflective materials with a pressure-velocity sensor (Microflown [8], or pu-sensor). Then, a series of repeated measurements are conducted to assess the uncertainty of EA method on both absorptive and reflective materials in the frequency region of 100 Hz to 1,500 Hz. To examine the validity of the mathematical-physical model described above and knowhow, tube methods of two kinds, at a laboratory and in situ, are conducted on the parts of the same materials. Then the agreement between methods is observed.