Several studies of benign breast lesions using methacran-fixed, paraffin-embedded tissues and cytological preparations have suggested that p53 accumulation in these lesions as detected by immunohistochemical (IHC) staining is rare to absent. As a result, several different investigators have suggested that p53 immunoreactivity in breast specimens infers a diagnosis of malignancy or may identify premalignant lesions. We immunostained 271 breast biopsy specimens from 271 patients with the monoclonal anti-p53 antibody BP-53-12 and found positive nuclear staining in seven of 23 malignant lesions (30%) and 39 of 248 benign biopsy specimens (16%). Of the benign lesions, 30% of fibroadenomas, nonpremalignant breast lesions, were positive. Long-term follow-up information was available on 48 patients with benign biopsy specimens and showed that 12% of those positive and 7% of those negative for p53 developed breast carcinoma. This difference was not significant (P > .2). We conclude that (1) p53 immunoreactivity in breast lesions should not be used as exclusive evidence of malignancy and (2) p53 immunoreactivity in benign breast lesions may not identify a subset of patients at increased risk for breast carcinoma.