OBJECTIVE: To compare the effect of perilaryngeal manual massage and traditional vocal training for professors with vocal complaints. METHOD: Forty-two university professors were randomized into 2 groups: perilaryngeal manual massage (G1) or vocal training (G2). Interventions included vocal and pain autoperception, vocal symptom historytaking, evaluation of the cervical muscle tension, and analyses of the percipient auditory and acoustics of the voice. RESULTS: The 2 groups showed a reduction in vocal symptoms. There was no difference between the interventions with respect to the partial and total scores of vocal self-assessment questionnaire or the acoustic analysis. G1 intragroup analysisshowed differences (p=0.013) in vocal autoperception and the acoustic parameter energy for glottic noise (p=0.003);G2 scoresshowed differences in the effects on daily communication (p=0.006), effects on emotion (p=0.007), activity limitation (p=0.036), total score (p=0.003), shimmer parameters (p=0.009), and glottal noise energy (p=0.000). G1 showed intragroup differences for pain autoperceptionas well as reducedcervical tension in relation to the discrete degree of dysphoniaand increased percentage of citizens with normal perception. G2 kept the result of the auditory percipient after analysis of the voice the intervention, and it did not demonstrate a difference in the tension evaluation. CONCLUSION: Both interventions improved the participants' well being and vocal quality;the speech therapist could decide which of the interventions to useafter observingthe complaints and vocal demands of the professors. Clinical Trial=3123