In 1996, a National Cancer Institute committee recommended four categories as uniform terminology for breast fine-needle aspirations (FNAs): benign, malignant, suspicious/probably malignant, and atypical/indeterminate. The latter is a controversial category. This study evaluates the usefulness of the atypical/indeterminate term, and examines sources of diagnostic equivocation in breast FNA. Eight hundred and twenty-two consecutive breast FNAs were previously classified as benign, malignant, suspicious, or unsatisfactory. Two hundred and thirteen (25.9%) cases had surgical follow-up and were classified as true positive (TP), false positive (FP), true negative (TN), false negative (FN), true suspicious (TS), or false suspicious (FS). Slides from FN, FP, TS, and FS were reviewed for interpretative error, poor clarity or preservation, obscuring material, sampling error, or insufficient malignant criteria. Cases were also evaluated as to whether classification as "atypical/indeterminate" would have improved patient care. There were 21/822 (2.6%) FN, 37/822 (4.5%) TS + FS, and 0 FP diagnoses. Seventy percent of suspicious diagnoses showed cancer on follow-up. The majority of FN and suspicious cases were due to sampling problems and insufficient criteria of malignancy. None were deemed more appropriately classified as "atypical/indeterminate" All required surgical confirmation for treatment. All equivocal breast diagnoses are due to similar problems. Splitting them into "suspicious/probably malignant" and "atypical/indeterminate" would not lower the biopsy rate. A simpler three-part terminology of benign, malignant, and suspicious/equivocal, without qualification of the latter favoring benign or malignant, would provide more effective communication and appropriate follow-up. Diagn. Cytopathol. 1999;21:217-222.