The p53 tumor suppressor gene is frequently mutated in breast cancer. Here, we used direct sequencing to screen the complete coding sequence of the p53 gene from 41 human breast cancer cell lines. We identified 32 cell lines (78%) with a p53 gene alteration that predicted a change in the encoded protein. Thirty-one of these mutations were accompanied by loss of the other p53 allele. All mutations but one were unique and 27 mutations had previously been identified in uncultured human cancers. Ten mutations were predicted to encode a truncated p53 protein and 22 missense mutations were identified. p53 transcript expression was analyzed by semi-quantitative RT-PCR and p53 protein expression was determined by Western blotting. Our analyses revealed three p53 expression patterns: wild-type p53 cell lines had normal transcript levels and low or no detectable protein expression; cell lines with a p53 truncating mutation had low transcript levels and low or no detectable protein expression; and cell lines with a p53 missense mutation had highly variable transcript and protein expression levels. As a whole, our data represent a p53 mutation profile in breast cancer cell lines, providing a model for structural, functional and pharmacological studies on p53 in human cancer.