Abstract Background The role of intraoperative neuromonitoring (IONM) of the recurrent laryngeal nerve (RLN) during thyroid surgery is still debatable. The aim of this meta-analysis was to evaluate the potential improvement of IONM versus RLN visualization alone (VA) in reducing the incidence of vocal cord palsy. Methods A literature search for studies comparing IONM versus VA during thyroidectomy was performed. Studies were reviewed for primary outcome measures: overall, transient, and permanent RLN palsy per nerve and per patients at risk; and for secondary outcome measures: operative time; overall, transient and permanent RLN palsy per nerve at low and high risk; and the results regarding assistance in RLN identification before visualization. Results Twenty studies comparing thyroidectomy with and without IONM were reviewed: three prospective, randomized trials, seven prospective trials, and ten retrospective, observational studies. Overall, 23,512 patients were included, with thyroidectomy performed using IONM compared with thyroidectomy by VA. The total number of nerves at risk was 35,513, with 24,038 nerves (67.7%) in the IONM group, compared with 11,475 nerves (32.3%) in the VA group. The rates of overall RLN palsy per nerve at risk were 3.47% in the IONM group and 3.67% in the VA group. The rates of transient RLN palsy per nerve at risk were 2.62% in the IONM group and 2.72% in the VA group. The rates of permanent RLN palsy per nerve at risk were 0.79% in the IONM group and 0.92% and in the VA group. None of these differences were statistically significant, and no other differences were found. Conclusions The current review with meta-analysis showed no statistically significant difference in the incidence of RLN palsy when using IONM versus VA during thyroidectomy. However, these results must be approached with caution, as they were mainly based on data coming from non–randomized observational studies. Further studies including high-quality multicenter, prospective, randomized trials based on strict criteria of standardization and subsequent clustered meta-analysis are required to verify the outcomes of interest.