The average frequency spacing 〈Δfmax〉 between adjacent maxima of the frequency response curve between two points in a room is determined by experiment, theory, and Monte Carlo computation. In earlier papers it had been shown that, above a certain critical frequency, 〈Δfmax〉 is reciprocally related to reverberation time and not dependent on other room characteristics—disproving a belief that 〈Δfmax〉 is a useful measure of the acoustical quality of rooms. Theory predicts〈Δfmax〉 = 3.91/T60, whereT60is the reverberation time. Monte Carlo computation gives 3.90/T60. Measurements in two very different rooms using a vacuum tube voltmeter for reading the sound pressure are in good agreement with these predictions. Measurements with a logarithmic level recorder give, as in earlier investigations, much larger values. The discrepancy cannot be explained by the level recorder's 0.5‐db quantization. This paper shows that: (1) The Monte Carlo method is a useful tool for solving complex problems in room acoustics. (2) Level recorder measurements of frequency (or space) irregularities of rooms must be taken with several grains of salt.