To evaluate intrinsic nodule features predictive of an inadequate report in fine-needle aspiration cytology (FNAC). Single-centre cross-sectional study. Between May 2005 and April 2011, 3279 ultrasonography-assisted FNACs were carried out and features of nodules recorded prospectively. Univariate logistic regression analyses were performed to estimate the association between nondiagnostic cytology and variables such as age, gender, single nodule, maximum nodule diameter and estimated volume. Inadequate or nondiagnostic samples were reported in 1195 FNACs. All diameters were found to be predictors of nondiagnostic cytology; estimated nodule volume, on the other hand, was not. Nodules with a diameter <10 mm were more frequently nondiagnostic (OR 1.65, 95% CI 1.40-1.94, P < 0.001). Neither micro- nor macrocalcification increased the risk of inadequacy. On the contrary, mixed lesions were more frequently diagnostic (OR 0.68, 95% CI 0.85-0.80, P < 0.001). Solid nodule aspiration was performed more easily on isoechogenic nodules (OR 0.64, 95% CI 0.54-0.77, P < 0.001); the same procedure was more cumbersome on hypoechogenic lesions (OR 1.87, 95% CI 1.62-2.16, P < 0.001). Increased vascularization did not cause a significant increase in the nondiagnostic results. Blurred margins increased the inadequacy rate (OR 1.45, 95% CI 1.24-1.69, P < 0.001), while presence of a hypoechogenic halo decreased it (OR 0.67, 95% CI 0.54-0.82, P < 0.001). Some ultrasonographic features suggestive of malignancy may be predictive of inadequate cytology. Patients must be notified that the FNA report may be nondiagnostic and that this represents a limitation of the technique related to the structure of lesions. © 2012 Blackwell Publishing Ltd.