Angiogenesis is essential for tumor growth and metastasis. There are conflicting reports as to whether microvessel density (IMD) in breast cancers is associated with prognosis. This could be due to the use of different antibodies to endothelial cell markers, variation in tissue pretreatment protocols, and nonstandardized counting methods. We have assessed the IMD in 106 breast carcinomas using a pan-endothelial marker, CD34, and a recently described mAb to CD105, which preferentially reacts with endothelial cell in angiogenic tissues. IMD values (separated as above or below median) for CD105 expression showed a statistically significant correlation with overall (P = 0.0029) and disease-free survival (P = 0.0362). In contrast, blood vessel counts using a panendothelial marker CD34 did not correlate with overall or disease-free survival (P = 0.2912 and P = 03153, respectively). When IMD values were subdivided into quartiles and assessed for their prognostic values, there was a statistically significant difference in the overall survival across CD105, but not CD34, values (P = 0.0017 and P = 0.7997, respectively) and also disease-free survival (P = 0.0431 and P = 0.5066, respectively). Further analysis of IMD values demonstrated that there were no deaths in the lowest quartile for CD105 and it differed from the other three quartiles. However, examination of clinical details of patients in the lowest quartile failed to reveal clustering of patients known to be associated with low-risk factors. Multivariate analysis confirmed that IMD values using CD105 were an independent prognostic factor. These results suggest that the ability to quantitatively distinguish between tumor neovascularization and preexisting vessels may be important in the assessment of tumor angiogenesis, but requires confirmation in a greater number of patients with a longer follow-up.