Fine-needle aspiration (FNA) is currently the most reliable diagnostic means in the clinical work-up of thyroid nodules. However, most of the available data on the diagnostic reliability of thyroid FNA derives from biopsies done just before surgery or from short-term observations, whereas data on the long-term reliability of benign FNA results, is extremely limited. Over the 17-year period between 1979 and 1996, thyroid FNA performed on 849 patients in our endocrine clinic. An initially benign result was reported in 578 patients for a total of 631 nodules. Mean follow-up period was 8.1 +/- 4.4 years (+/-standard deviation [SD]; median, 6.9 years). In order to ensure the completeness of our follow-up data, we supplemented our own patients' data with data from the Israel Cancer Registry for documentation of thyroid malignancy. Sixty-six of the patients with an initially benign FNA diagnosis were rebiopsied during follow-up. Five patients (0.85%) of all those with an initial benign FNA diagnosis, were subsequently found to have thyroid malignancy diagnosed 6 months or more after the initial evaluation. Three of the newly diagnosed malignancies were follicular and two were papillary carcinomas. Three of the patients had elements of being at high-risk: previous head irradiation, previous thyroid surgery with an occult cancer, and a growing goiter, respectively. Only 1 of 35 patients who had more than one benign FNA results was subsequently diagnosed with thyroid malignancy (follicular carcinoma). These results indicate that the rate of subsequent thyroid malignancies in patients with an initial benign FNA diagnosis is low, and thus benign thyroid FNA results provide a high level of long-term assurance. Still, repeating FNA is warranted in patients with longstanding thyroid nodules, particularly if at increased risk for cancer, or if a goiter is found to have changed its morphological characteristics over time.