In experimental conditions the influence of 8 ratios "useful signal-masking noise" of 3 types of production noise (impulse, continuous and interrupted) is studied as well as continuous (white) noise as control on the perception of an audiometric test by 60 individuals with normal hearing. The test is understandable phonetically with difficulties and balanced at speech. The test intensity is Lequiv.70–75 dB/A. The most expressed masking effect of continuous (white) noise, followed by the impulse and constant production noise is found. The lowest masking effect is found out at interrupted noise. This is explained by the acoustic characteristics of exposed types of noise. The masking effect of noise is limited to a high degree by the parameter "signal – noise" ratio, which is confirmed by the found negative correlation relationship with percentage speech intelligibility. The critical ratios "signal – noise", characterized with sharp decline of the level of discrimination of speaking signal and deterioration of the effectiveness of speech communication are established. In this paper the problem for conforming the speech communication both with the intensity and acoustic characteristics of masking noise and with the "signal – noise" ratio is discussed. Besides this in the critical "useful signal – masking noise" ratios better individual achievements could be a criterion for professional selection of candidates for operators.

        Key words: noise, speech communication, masking effect