The aim of this study was to investigate whether tumour hypoxia and/or vascular hot spots promote the development of metastatic disease. The D-12 human melanoma xenograft line was used as a tumour model. Hypoxia and vascular hot spots were detected by immunohistochemistry using pimonidazole as a hypoxia marker and anti-CD31 antibody to visualize endothelial cells. Vascular hot spots were found to be induced in hypoxic foci, owing to hypoxia-induced up-regulation of angiogenesis stimulatory factors. This effect was mediated by interleukin 8 and possibly also by vascular endothelial growth factor. Interleukin 8 positive foci showed a high degree of co-localization with hypoxic foci, as revealed by immunohistochemistry. The incidence of spontaneous pulmonary metastases was associated with the density of hypoxic foci, the density of interleukin 8 positive foci and the density of vascular hot spots in the primary tumour. Treatment with neutralizing antibody against interleukin 8 and/or vascular endothelial growth factor resulted in hypoxia-induced necrosis rather than hypoxia-induced vascular hot spots and inhibited metastasis. Our study suggests a cause-effect relationship between hypoxia and metastasis in cancer and hence an elevated probability of metastatic disease in patients having primary tumours characterized by high densities of hypoxic foci and vascular hot spots. Most human tumours develop a pathophysiological microenvironment during growth, characterized by an irregular microvascular network and regions of chronically and transiently hypoxic cells (Vaupel et al, 1989). Side by side with the generation of an abnormal microenvironment, tumours gradually acquire aggressive phenotypic traits with time, a process termed malignant progression (Nowell, 1986). The final stage of the malignant progression is the development of cell variants showing invasive growth in surrounding normal tissues and metastatic spread to regional and distant organ sites. There is significant evidence that the microenvironment may accelerate the malignant progression of tumours and promote the development of metastatic disease (Rofstad, 2000). Thus, some tumour cells exposed to hypoxia in vitro show increased lung colonization efficiency after intravenous inoculation in mice (Young et al, 1988;Rofstad and Danielsen, 1999). Moreover, clinical studies have demonstrated associations between incidence of metastases and extent of hypoxia in the primary tumour in soft tissue sarcoma (Brizel et al, 1996) and cervical carcinoma (Höckel et al, 1996;. These clinical studies, however, do not necessarily implicate a cause -effect relationship between hypoxia and metastasis. An alternative interpretation is that poor oxygenation is a secondary effect of tumour aggressiveness, i.e. the most aggressive cell phenotypes develop the most hypoxic primary tumours. There is also significant evidence that tumour hypoxia may be of limited significance in cancer metastasis (Rofstad, 2000). Thus, clinical studies involving several histological types of cancer...