Ultrasonography-guided fine needle aspiration (FNA) and fine needle non-aspiration (FNNA) biopsy were performed consecutively on 104 patients (mean age 50 +/- 15 years) affected by multinodular and uninodular goiter. Both techniques were executed on the same patients in the same clinical session beginning with the first (FNA) on half of the patients (randomly selected) and vice versa. The cytological findings on cell samples were divided into four groups: colloidal (benign), follicular (suspicious), malignant, and inadequate. The overall cytologic findings obtained with the two techniques were as follows (FNA % vs FNNA %): inadequate, 16.3% vs 5.8%; colloidal, 69.2% vs 76.9%; follicular, 9.6% vs 10.5%; and malignant, 4.8% vs 6.7%. A statistically significant difference between FNA and FNNA cytology was found only on the number of inadequate results (p = 0.015). Interestingly, the frequency of inadequate specimens for FNNA showed a significantly different distribution depending on the sequence (first or second) in which the technique was executed. In conclusion, FNA and FNNA are useful and cost-effective techniques for the pre-operative assessment of patients with thyroid nodules. However, due probably to its minimally invasive procedure, FNNA produces specimens of better quality and reduces inadequate results. For these reasons FNNA should be preferable to FNA for the cytological evaluation of thyroid nodules.