Small-bore (22- or 23-gauge) needles were used to aspirate 458 lung masses. Sensitivity for the detection of malignancy by this method was 96.6% (312 of 323 patients); accuracy was 98.7%. Surgical confirmation was available for comparison in over half of patients with malignancy. Histologic reclassification of malignancy occurred in only 6.1% of patients, with significant misclassification (misdiagnosis of small cell carcinoma) occurring in only two instances. One hundred thirteen of 117 nonmalignant conditions were properly categorized, with an overall specificity of 96.6%. No major complication occurred. While several recent studies have stressed the advantage of using larger needles, to overcome the limitation of smaller aspiration needles that provide only cytologic material, small-needle aspiration appears to be a safe, reliable, and accurate means for diagnosing lung lesions.