BACKGROUND. Endoscopic ultrasound (EUS) is a powerful imaging modality to identify and determine the extent of a lesion. In addition, EUS is superior to a computed tomography scan in detecting lesions < 3 cm. The objective of the current study was to determine whether small lesions (≤ 25 mm) affected the specimen adequacy and the diagnostic accuracy for lesions aspirated under EUS guidance. METHODS. In the current study, 209 consecutive EUS-guided fine-needle aspiration biopsy (EUS-FNAB) samples ≤ 25 mm (100 samples) or > 25 mm (109 samples) as determined by EUS were obtained from 151 patients with a mean age of 62 years (range, 39-94 years). A cytopathologist present in the endoscopy suite determined specimen adequacy. Yield of adequate samples for diagnosis, number of passes, and operating characteristics of EUS-FNAB for small (≤ 25 mm) and large lesions (> 25 mm) were compared. RESULTS. The overall yield of obtaining adequate samples for diagnosis was 96% (201 of 209). There was no difference noted with regard to the yield of obtaining samples (96% vs. 96%) from small or large lesions. A mean of 2.5 passes (range, 1-9 passes) was needed to obtain adequate samples from lesions ≤ 25 mm, whereas a mean of 4.5 passes (range, 1-11 passes) was needed to obtain adequate samples from lesions > 25 mm. The sensitivity (96% vs. 96%), specificity (100% vs. 100%), and diagnostic accuracy (98% vs. 97%) for EUS-FNAB were comparable whether the lesion was ≤ 25 mm or > 25 mm. CONCLUSIONS. EUS-FNAB was a highly effective technique to obtain samples from small (≤ 25 mm) and large (> 25 mm) lesions without affecting the sensitivity, specificity, or diagnostic accuracy. © 2004 American Cancer Society.