Introduction: Continuous intraoperative neuromonitoring (kontIONM) and the provision of relevant information such as moment, origin and prognosis of nerve function impairment during thyroid resection have been tested. Methods: Between 2009 and 2011, 667 patients were operated for thyroid pathology by applying kontIONM (tube electrode, vagal probe V3, ISIS; Fa. Inomed, Emmendingen, Germany). Vocal cord function was examined laryngoscopically on the 2nd postoperative day. Palsies were diagnosed in 34 patients. Complete kontIONM signals were filed during the operation. Loss of signal (LOS), defined as amplitude reduction < 100 mu V, and signal delay > 10% were attributed to thyroid dissection. Results: A LOS of 17.6% (6/34) developed already at the moment of thyroid lobe luxation, that is, prior to a dissection for recurrent laryngeal nerve (NLR) identification. An LOS of 67.6% (23/34) appeared during NLR preparation in the vicinity of the Berry ligament. Thus, 85.3% of all vocal cord palsies were recognised intraoperatively. For four patients signal delay > 10% could be observed in the analysis of the postoperative signal but not during the operation. One case was not associated with any of these signal changes. Conclusion: In the majority of cases, signal loss and reduction of amplitude < 100 mu V are reliable parameters of post-operative vocal cord palsy. Traction and distension of the nerve seems to be the most important cause of nerve damage. An immediate revision of the last step of the surgical procedure, if required, is the essential advantage of this method to avoid irreversible nerve damage. For a minor part of the cases, vocal cord palsies are characterised intraoperatively by an extended delay of the signal.