Objective To study the correlation between preoperative serum thyroid stimulating hormone (TSH) level and the malignant risk of thyroid nodules. Methods This study retrospectively analyzed the medical records of 959 patients with thyroid nodules. Ststistical analysis was conducted by SPSS 17.0 software. Results There were 959 patients with thyroid nodules, of which 746 cases were benign, and 213 cases were diagnosed as thyroid papillary carcinoma (PTC). The preoperative TSH level of PTC patients was higher than that with benign nodules [(2.32±1.65) mU/L vs. (1.76±1.20) mU/L, P<0.001]. Moreover, the higher preoperative TSH level was, the higher risk of diagnosed as PTC would be. There was no correlation between the preoperative TSH level and tumor diameter, number of lesions, and lymph nodes metastasis in PTC (P>0.05). Logistic regression analysis showed that, the preoperative TSH level was an independent risk factor for PTC [OR=1.315, 95% CI was (1.171, 1.477), P<0.001]. The best critical value of TSH in the PTC diagnosis was 1.575 mU/L. At this point, the sensitivity was 62.0%, the specificity was 53.4%, and the area under the receiver operating characteristic (ROC) curve was 0.602 (P<0.001). Conclusion There is a certain correlation between preoperative TSH level and malignant risk of thyroid nodules, and the risk increases with the raise of preoperative TSH level.