Fine-needle aspiration (FNA) is a diagnostic modality that continues to improve in accuracy as training and experience accumulate. With increasing operator expertise and improved localization techniques, greater numbers of patients are able to benefit from FNAs performed on sites that are otherwise difficult or dangerous to reach by conventional surgery. We present a retrospective review of a 2-yr experience with radiologically-guided deep-seated FNA. In 115 cases involving transthoracic and transabdominal sites, we achieved the following overall figures: 91.9% sensitivity, 100% specificity, 93.9% diagnostic accuracy, 100% positive predictive value, and 80.6% negative predictive value. Our results are compared to those in other series. When properly applied, FNA of deep-seated lesions through image guidance is equivalent to tissue diagnosis obtained by laparotomy or surgical procedures. The benefits of FNA with or without core biopsy vs. scalpel biopsy are readily apparent when one considers the morbidity, cost, turnaround time, and trauma to the patient.