Methods for measuring masking level differences (MLDs) at 500 Hz and for spondees were used with 290 subjects: 50 persons with normal hearing and 240 patients with various diseases. Of particular interest was whether techniques currently in clinical use could be used with ease, dispatch, and profit in determining MLD size. The methods selected, which were variations of Békésy audiometry and speech reception threshold procedures, proved clinically feasible. Results revealed differences in behavior from one group of subjects to another. Although MLDs were not affected by cortical lesions, they were very often abnormally small for patients with eight-nerve tumor, Meniére's disease, or multiple sclerosis. The high incidence of abnormally small MLDs in populations with normal sensitivity to pure tones and speech but with evidence of subcortical central lesions, such as patients with multiple sclerosis, suggests that the MLD tasks can be of diagnostic value in detecting retrocochlear lesions. However, in persons with hearing loss or significant interaural differences in threshold sensitivity, or both, the MLD tests are not always reliable in differentiating cochlear from retrocochlear disease.