Purpose: To systematically appraise and summarize the available evidence about the diagnostic accuracy of ultrasound-guided fine needle aspiration biopsy (USFNA) for thyroid malignancy, and to explore the integration of these estimates with the probability of thyroid malignancy before USFNA. Methods: A comprehensive search of multiple databases from each database’s inception to August 2014 was performed. Eligible studies included those that evaluated patients with thyroid nodules who underwent USFNA and subsequent evaluation by histopathology or long-term follow-up. Results: We identified 32 studies at moderate risk of bias evaluating the USFNA diagnostic characteristics for the diagnosis of thyroid malignancy. Results were imprecise and inconsistent across trials. The pooled likelihood ratio (LR) of thyroid malignancy for a benign USFNA result was 0.09 (95 % CI 0.06, 0.14; I2 = 33 %), whereas the pooled LR for a malignant result was 197 (95 % CI, 68, 569; I2 = 77 %). In the case of a suspicious for follicular neoplasm result, the pooled LR for malignancy was 0.6 (95 % CI, 0.4, 1.0; I2 = 84 %) and 8.3 (95 % CI, 3.6, 19.2; I2 = 89) for a result of suspicious for malignancy. Conclusion: The available evidence regarding the diagnostic accuracy of USFNA warrants only limited confidence due to risk of bias, imprecision, and inconsistency. However, some USFNA results (benign, malignant) are likely very helpful, by significantly changing the pre-test probability of thyroid cancer.