Telomere shortening leads to genomic instability and has been correlated with poor outcome in several types of cancer. A recently described, robust titration assay was used to quantify telomere DNA content in frozen and paraffin-embedded specimens of 49 invasive human breast carcinomas, including tumors with normal or abnormal contents of genomic DNA, which produced regional, distant, or local disease. Telomere DNA contents ranged from 53% to 370% of the content in a reference DNA purified from normal placenta. Tumors were divided into three groups of approximately equal size based on increasing telomere DNA content. All of 16 tumors in the group with the least telomere DNA (Group I), were aneuploid compared to 9/17 tumors in the group with the most telomere DNA (Group III). The Chi-square test for trend indicated that tumors with the least telomere DNA were significantly more likely to be aneuploid than tumors with the most telomere DNA (p < 0.002). Twelve of 14 tumors in Group I also produced metastatic disease compared to 8/15 tumors in Group III. The Fischer Exact Test indicated that tumors with the least telomere DNA were significantly more likely to be metastatic than tumors with the most telomere DNA (p < 0.05). There was no association between telomere DNA content and patients' age, tumors' size, grade, stage, or fraction of cells in S-phase. The correlation of reduced telomere DNA content with aneuploidy and metastasis, both of which are associated with poor outcome in invasive breast carcinoma, implies that telomere DNA content also could have prognostic value.