Neuromonitoring in thyroid surgery has been employed to make nerve identification easier and decrease the rates of laryngeal nerve injuries. Several individual randomized controlled trials (RCTs) have been published, which did not identify statistical differences in the rates of recurrent laryngeal nerve (RLN) or external branch of the superior laryngeal nerve (EBSLN) injuries. The objective of this report is to perform meta-analysis of the combined results of individual studies to measure the frequency of RLN and EBSLN injuries in patients who underwent thyroidectomy with routine neuromonitoring in comparison with common practice of search and identification. RCTs comparing routine neuromonitoring versus no use in patients who underwent elective partial or total thyroidectomy were evaluated. Outcomes measured were temporary and definitive palsy of the RLN and EBSLN. A systematic review and meta-analysis was done using random effects model. GRADE was used to classify quality of evidence. Six studies with 1,602 patients and 3,064 nerves at risk were identified. Methodological quality assessment showed high risk of bias in most items. Funnel plot did not reveal publication bias. The risk difference for temporary RLN palsy, definitive RLN palsy, temporary EBSLN palsy, and definitive EBSLN palsy were -2% (95% confidence interval -5.1 to 1); 0% (-1 to 1); -9% (-15 to -2) and -1% (-4 to 2), respectively. Quality was rated low or very low in most outcomes due to methodological flaws. Meta-analysis did not demonstrate a statistically significant decrease in the risk of temporary or definitive RLN injury and definitive EBSLN injury with the use of neuromonitoring. The neuromonitoring group had a statistically significant decrease in the risk of temporary EBSLN injury.