This study aimed to investigate the diagnostic value of doing a second ultrasound-guided fine-needle aspiration (US-FNA) for thyroid nodules of different sizes that could not be diagnosed by the first US-FNA.One hundred and forty-three patients (162 nodules) were diagnosed with suspected malignant thyroid nodules in a routine ultrasound examination, but since the diagnosis could not be confirmed by the cytology of the samples collected in the first US-FNA, the patients underwent US-FNA again 3 months later. The ultrasound results, cytology results, and postoperative pathology of these nodules were collected. The nodules were divided into three groups according to the largest diameter (L) of the thyroid nodules: Group 1, L < 0.5 cm, 26 nodules; Group 2, L = 0.5-1.0 cm, 76 nodules; and Group 3, L > 1.0 cm, 60 nodules.In the second US-FNA, the overall diagnosis rate of the 162 thyroid nodules that could not be given a definitive diagnosis by the first US-FNA was 51.8% (84/162). The definitive diagnosis rates of the nodules in Groups 1, 2, and 3 were 30.8% (8/26), 67.1% (51/76), and 41.7% (25/60), respectively. The diagnosis rate was the highest in Group 2, and the differences between this group and the other two groups were statistically significant (χ2  = 10.489, 8.801, p < 0.05 for both). The diagnostic accuracy rates of Groups 1, 2, and 3 were 100% (8/8), 96.1% (49/51), and 92% (23/25), respectively.Second US-FNA is highly recommended for such nodules.© 2022 Wiley Periodicals LLC.