Cough is a physiological gesture witch requires a perfect laryngeal competence. Tussometry was described as a reliable and reproducible method of analysis based on the measure of oral air flow rate during cough. This test allows to quantify the efficiency of laryngeal occlusion in normal situation and in case of unilateral laryngeal paralysis. The measure of the time lag between the beginning of the cough and the maximal air flow rate (peak value time) is the most reliable parameter (8). Present study is designed to asses the reproducibility of tussometry realized on a vocal analysis workstation EVA II and to test the efficiency of intracordal injection of autologus fat in case of unilateral laryngeal paralysis following thoracotomy. 10 control subjects were recorded at two moments. 6 patients presenting unilateral laryngeal paralysis following thoracic surgery were recorded before and after injection of autologus fat in the paralysed vocal cord. Reproducibility of tussometry is good (r = 0.96). The mean peak value time is significantly improved (p = 0.048). In conclusion, tussometry is a reliable routine test on EVA III workstation. Intracordal autologus fat injection improves occlusive function of larynx during cough in case of unilateral paralysis.