Although several tumour types express both AT1 and AT2 angiotensin II receptors, and angiotensin II stimulates cell proliferation, angiotensin-converting enzyme inhibitors and angiotensin receptor blockers are not effective anti-cancer agents. Development of a biologically active monoclonal antibody (6313/G2) against the AT1 receptor prompted the testing of a recombinant short-chain variable fragment form (R6313/G2) against breast cancer cells in vitro and in vivo. Cell lines MCF-7, MDA-MB-231 and T47D all expressed both receptor subtypes. In vitro, R6313/G2 suppressed cell proliferation in the presence of 100 nM angiotensin II, with IC50s of 30 nM, 153 nM and 2.8 microM for the three cell types respectively; in contrast, the AT1 receptor blocker losartan was effective only in T47D cells, at 25 microM. Studies on MCF-7 and T47D cells showed R6313/G2 also opposed the angiotensin II-induced inhibition of caspase-3/7 activity. In vivo, hollow fibres containing the cell lines were implanted in nu/nu balb-c mice at two sites, s.c. and i.p. Treatments of R6313/G2 at 2.5 nmol/kg and 25 nmol/kg twice per day for 7 days dose dependently reduced cell numbers for all three cell lines, but here MCF-7 cells responded most sensitively and MDA-MB-231 cells least. Although T47D cells were refractory at the s.c. site, growth was inhibited at the i.p. location, and otherwise results were similar at the two sites. In xenografts, MCF-7 cell tumours were dose dependently reduced by R6313/G2, and 13 and 27 nmol/kg R6313/G2 twice/day gave means of 74 and 76% tumour regression after 7 days. The data suggest that the anti-cancer action of R6313/G2 is considerably more effective than AT1 antagonists.