A series of experiments were performed to compare the efficiency of different diffusing devices in producing a completely random sound field in a reverberation chamber and the effectiveness of some of the methods for judging whether a sound field is completely random. An additional experiment was performed to compare two types of test signals, viz., warble tone and the same nominal band width of random noise. Results of these measurements indicate the general superiority of the rotating vane over other diffusing devices and superiority of random noise over warble tone in producing a smooth decay curve. One of the significant conclusions is that even when the diffusion is reasonably adequate the measured absorption coefficient of a standard sample appears to vary with the chamber absorption especially at the high frequencies.