At the Maharaj Nakorn Chiang Mai Hospital, Chiang Mai, Thailand. 2,375 cases of breast lesions were sampled by fine-needle aspiration (FNA) from 1994-1999. Cytologic diagnosis were: benign (48%), suspicious for malignancy (5%), malignant (15%), and unsatisfactory (32%). Comparison with histology was possible in 721 cases. The diagnosis obtained by FNA showed a sensitivity of 84.4%, specificity of 99.5%, positive predictive rate of 16.7%, false-positive rate of 0.5%, and overall diagnostic accuracy of 91.3%. We conclude that, in experienced hands, FNA of breast masses is reliable for diagnosis. Assessment of sample at the times of aspiration can reduce the number of inadequate specimens to near zero. Correlation of FNA results with clinical and radiologic findings can identify false-negatives and false-positives, ensuring optimal patient management. Many centers now recommend needle core biopsy instead of FNA. For regions such as ours, the added cost of this would make it available to many patients, which could delay a diagnosis of breast cancer. We advocate keeping FNA as a first-line diagnostic procedure, at least in areas under economic restrictions, in order to maximise the availability of health care to women with breast cancer.