In an attempt to confirm earlier results with a group of mentally retarded females, 12 mentally retarded institutionalized adults (8 male, 4 female) were trained to either reduce (Loud group) or increase (Soft group) their voice volumes with a combination of visual feedback and token reinforcement. The feedback unit provided a binary light on-off signal. Colored lights were used for loud (red), soft (blue), and normal (green) voice volumes, and a digital summary was provided of the total time each light was on. Tokens were given for each trial of 15 sentences with a green light time-on of 70% or better. One-half of the sample received feedback only; the other half was given feedback plus tokens. Results produced a significant improvement in control of voice volume in Loud and Soft groups as measured by the differences in pre- and post- baseline decibel readings (5, 12, and 19 days post-training), cumulative tools of the light readouts, and caseworkers' ratings of each participant. The results confirmed findings on the effectiveness of visual feedback. It is suggested that the use of visual feedback as a training device can be used not only to address voice intensity disorders, but other voice disorders, as well, in this population. (Author/CL)