Background. The purpose of this study was to determine whether ultrasonography (US) improves the accuracy and reduces the rate of nondiagnostic fine-needle aspiration biopsy (FNAB) of thyroid nodules. Methods. A review of 536 consecutive patients evaluated for nodular thyroid disease from 1990 to 2001 was completed to determine the results for US vs palpation-guided FNAB. Results. FNAB was used to evaluate 458 patients. US-guided FNAB was performed in 66 (12%) patients: 48 with a nonpalpable nodule, 14 with nondiagnostic standard FNAB, and 4 with a palpable nodule. US-guided FNAB was nondiagnostic in 15 (23%) patients. There were no false-positive or false-negative results. Standard FNAB was performed in 407 patients, 57 (14%) of whom had a nondiagnostic result. There were 2 (3%) false-positive and 3 (1.6%) false-negative results. In 14 patients with a nondiagnostic standard FNAB, US-guided FNAB yielded an adequate specimen in 7 (50%). Nodules evaluated by standard FNAB were 4.1 ± 0.1 cm (mean ± SEM) in size compared with 2.5 ± 0.1 cm for nodules evaluated by US-guided FNAB (P <.05). Conclusions. US improves the diagnostic yield in selected patients with nondiagnostic standard FNAB. The higher frequency of nondiagnostic US-guided FNAB was related to its selective use in patients with smaller nodules. Surgery 2002;132:648-54.