In the acoustical treatment of rooms, an accurate knowledge of sound absorption coefficients is necessary. Test data on Type B Acousti‐Celotex from seven different laboratories is presented showing a variation from .40 to .47 in values obtained by the reverberation method and from .23 to .40 by the “tube” method. The values .23, .40 and .47 were all obtained on the same samples shipped to the different laboratories. The desirability of standardizing test methods is apparent.To determine which absorption values should be used in engineering work, a study was made of a large number of completed installations. A small 512 cycle organ pipe was calibrated for acoustic output both at the University of Illinois and at the Bureau of Standards. The 3A Audiometer was used to determine the amount of noise, if any, in the rooms and it was found possible to correct the measurements for the threshold shift caused by the noise. From the data, 45 installations are selected as typical. Using .47 as the coefficient, the total absorption in each room was calculated. In 24 of the rooms the test absorption was greater than the calculated absorption by an average of 9.7%. In 21 of the rooms the test absorption was less than the calculated absorption by an average of 12.1%.