Background: When sonographic suspicious malignant thyroid nodules initially proven to be inconsistent by aspiration, repeat fine-needle aspiration cytology (FNAC) is commonly recommended. Although FNAC is simple and reliable, it is also an invasive operation with inevitable false negative results. To determine which sonographic features and histogram parameters based on gray-scale ultrasonic images need repeat aspiration.Methods: All of the nodules were suspicious for malignancy by preoperative ultrasound (US), and their initial FNAC findings indicated Bethesda category II, or III. Histogram analyses were performed for grayscale ultrasonic images. Mean, variance, skewness, kurtosis, percentiles (1st, 10th, 50th, 90th, 99th), and sonographic features were compared between pathological benign nodules and malignant ones. Thereafter, receiver operator characteristic (ROC) curves were analyzed for the valuable indicators.Results: One hundred and nineteen consecutive patients with 123 histopathologically diagnosed focal thyroid nodules were included in the cohort. The factors associated with malignancy were taller-thanwide shape [odds ratio (OR) =15.165; 95% confidence interval (CI): 3.157-72.854], irregular margins (OR =11.492; 95% CI: 1.747-75.573), microcalcifications (OR =5.107; 95% CI: 1.455-17.927) and skewness (OR =25.800; 95% CI: 1.034-76.422). The skewness of malignant thyroid nodules is higher than that of benign thyroid nodules and has an area under the curve (AUC) of 0.776 in the diagnosis.Conclusions: For thyroid nodules suspicious on US but not on cytology, skewness, together with tallerthan-wide shape, irregular margins and microcalcifications might be a promising marker for clinicians to perform repetitive FNACs.