Background To evaluate the diagnostic performance of elastography in the differentiating the malignancy from thyroid nodule. Methods The literature published in English or Chinese prior to December 2019 which diagnosed by elastography in thyroid nodules were reviewed and summarized. According the inclusion and exclusion criteria, the elegant theses were reserved. Quality of each study was evaluated by the qu ality a ssessment of d iagnostic a ccuracy s tudies tool (QUADAS). Comprehensive Meta-Analysis Version 2.0 was used to describe primary results and explore homogeneity. Publication bias analysis was evaluated by Review Manager 5.3. Meta-Disc version 1.4 was applied to explore threshold effect, and pool sensitivity and specificity. A Summary ROC curve was constructed to calculate the area under the curve. Results As the heterogeneity was mild, we performance on random effect model. The publication bias was negligible. The elastography showed high accuracy in the diagnosis of malignant. The pooled sensitivity and specificity with corresponding 95% confidence interval were 0.80 (0.78, 0.82) and 0.80 (0.79, 0.81) respectively. AUC of sROC was 0.9087. Sensitivity analysis demonstrated that the pooled estimates were stable and reliable. Conclusions The elastography has obvious sensitivity and specificity in differentiate malignant from the thyroid nodule. Several different types of elastography could achieve the same aim, but the cut-off value is associate with the purpose of the screen program and the subtypes of the technique.