Fifty-seven patients with large, oestrogen receptor (ER)-positive breast cancers were treated with a primary course of tamoxifen prior to surgery. Tumour tissue was assessed in terms of tumour vascularity before and after treatment and changes correlated with response. A significant association was found between change in tumour vascularity and response, with microvessel counts decreasing in responding tumours (p=0.0064) and tending to increase in non-responding tumours (p=0.036). Thus successful treatment with tamoxifen resulted in reduced tumour vascularity. To determine timing of these changes more accurately, mice bearing xenografts of ER-positive ZR-75-1 and ER-negative MDA-MB-231 breast cancer cell lines were treated with tamoxifen and assessed for changes in tumour size and vascularity. After 2 days, there was evidence of reduced microvessel counts in ZR-75-1, but not MDA-MB-231 tumours. Clear evidence of tumour regression in ZR-75-1 tumours only became apparent after 4 days. These results suggested that changes in vascularity precede tumour regression. Forty-two patients were recruited into a prospective study in which patients were treated with primary tamoxifen. Tumour tissue in the form of core biopsies was available prior to and after fourteen days of treatment. Patients underwent definitive surgery at the end of three months. Microvessel counts were compared at the three time points. The only significant difference identified between non-responding and responding tumours was increased microvessel counts in responding tumours following two weeks of treatment (p=0.01). This may be due to local stimulation of angiogenesis in areas of hypoxia. Response to primary treatment with tamoxifen was associated with reduced tumour vascularity, and xenograft data suggested that such changes occur prior to tumour regression. However, problems with reproducibility of microvessel counts in small tumour biopsies may limit their role as an early predictor of sensitivity to endocrine therapy.