An experimental investigation was made of the behavior of various model rocket chambers driven acoustically by cold air flowing through a porous plate at the head end. Analysis showed that bulk losses and viscous losses were negligible as compared to losses through the sonic nozzle. Nozzle losses were found to vary linearly with J (nozzle-throat/grain-port-area ratio) in the range tested (O<J<0.4). Results are compared with zero-length “linear-entrance-area” sonic-nozzle theory and also with linear acoustic theory for a piston-driven chamber having zero net flow. Three separate experimental methods were used to determine the nozzle attenuation constants. The steady-state technique was the only method that was satisfactory through the entire range of practical interest. The steady-state “decay-rate method” was limited to J's of less than 0.14. The pulse technique was not satisfactory. The tests show conclusively that the change in nozzle radiation losses during burning (caused by the variation of “J”) drastically alters the acoustic response of rocket chambers to a given excitation.