1. We investigated why angiotensin (Ang) I and II induce vasoconstriction with similar potencies, although Ang I-II conversion is limited. 2. Construction of concentration-response curves to Ang I and II in porcine femoral arteries, in the absence or presence of the AT(1) or AT(2) receptor antagonists irbesartan and PD123319, revealed that the approximately 2 fold difference in potency between Ang I and II was not due to stimulation of different AT receptor populations by exogenous and locally generated Ang II. 3. Measurement of Ang I and II and their metabolites at the time of vasoconstriction confirmed that, at equimolar application of Ang I and II, bath fluid Ang II during Ang I was approximately 18 times lower than during Ang II and that Ang II was by far the most important metabolite of Ang I. Tissue Ang II was 2.9+/-1.5% and 12.2+/-2.4% of the corresponding Ang I and II bath fluid levels, and was not affected by irbesartan or PD123319, suggesting that it was located extracellularly. 4. Since approximately 15% of tissue weight consists of interstitial fluid, it can be calculated that interstitial Ang II levels during Ang II resemble bath fluid Ang II levels, whereas during Ang I they are 8.8 - 27 fold higher. Consequently at equimolar application of Ang I and II, the interstitial Ang II levels differ only 2 - 4 fold. 5. Interstitial, rather than circulating Ang II determines vasoconstriction. Arterial Ang I, resulting in high interstitial Ang II levels via its local conversion by ACE, may be of greater physiological importance than arterial Ang II.