Any impairment of the audio-phonatory control by background noise is followed by an increase of the intensity level and the pitch level of the speaking voice (Lombard-Reflex, 1911) thus increasing vocal strain. Therefore, it might be expected that persons reacting to noise with very marked voice changes are more liable to develop dysphonia. In 22 singers, 35 normal persons and 22 patients with hyperfunctional dysphonia both ears were masked gradually with a white noise. The change of the mean intensity level and of the mean pitch level of the speaking voice were measured objectively with a special fundamental frequency analyzer (Fedders and Schultz-Coulon, 1975). Results show, that the increase of intensity is almost the same in all subjects, whereas the elevation of the mean pitch level differs significantly: trained voices (singers) react with the smallest, and dysphonic patients with the greatest pitch increment. The following conclusions are made: 1. Extreme increments in pitch level are a much more important etiological factor of dysphonia than intensity increments. 2. Voice therapy and voice training may have an favourable effect on the Lombard-Reflex (probably by improvement of the kinesthetic control-mechanism), so that the speaking voice in a noisy environment is raised less, thus diminshing vocal strain. Therefore, the measurement of pitch changes during binaural masking can give important informations for the diagnosis, therapy and prophylaxis of dysphonia.