Objective To investigate the role of preoperative serum thyroid stimulating hormone (TSH) in evaluating differentiated thyroid cancer (DTC). Methods A total of 551 patients with thyroid nodules met the study criteria, who got treatment in the First Affiliated Hospital of Jiamusi University between Aug. 2017 and Dec. 2017. And the patients were divided into DTC group (n=110) and benign group (n=441) according to the postoperative pathological results. The difference of serum TSH level between the 2 groups was compared and then explored the diagnostic significance of serum TSH level, thyroid imaging report and data system (TI-RADS), and serum TSH combined withTI-RADS. Results The serum TSH level was higher in the DTC group than that of the benign group (Z=5.198, P<0.05). The sensitivity of preoperative serum TSH level in the diagnosis of DTC was 80.9% (89/110), the specificity was 74.4% (328/441), and the area under receiver operating characteristic (ROC) curve was 0.660 ［95%CI was (0.602, 0.719), P<0.05］. The sensitivity of TI-RADS in the diagnosis of DTC was 82.7% (91/110), the specificity was 73.5% (324/441), and the area under the ROC curve was 0.772 ［95%CI was (0.711, 0.823), P<0.05］. The sensitivity of preoperative serum TSH combined with TI-RADS in the diagnosis of DTC was 91.8% (101/110), the specificity was 87.5% (386/441), and the area under the ROC curve was 0.831 ［95%CI was (0.786, 0.876), P<0.05］. Conclusions Preoperative serum TSH level may be a potential risk factor for the occurrence of DTC. Preoperative serum TSH level combined with TI-RADS classification can improve the accuracy of preoperative diagnosis of DTC, not only to reduce the misdiagnosis of thyroid cancer, but also can avoid excessive treatment.