The classical model of coronary physiology implies the presence of maximal microcirculatory vasodilation during myocardial ischemia. However, Doppler monitoring of coronary blood flow (CBF) documented severe microcirculatory vasoconstriction during pacing-induced ischemia in patients with coronary artery disease. This study investigates the mechanisms that underlie this paradoxical behavior in nine patients with stable angina and single-vessel coronary disease who were candidates for stenting. While transstenotic pressures were continuously monitored, input CBF (in ml/min) to the poststenotic myocardium was measured by Doppler catheter and angiographic cross-sectional area. Simultaneously, specific myocardial blood flow (MBF, in ml.min(-1).g(-1)) was measured by 133Xe washout. Perfused tissue mass was calculated as CBF/MBF. Measurements were obtained at baseline, during pacing-induced ischemia, and after stenting. CBF and distal coronary pressure values were also measured during pacing with intracoronary adenosine administration. During pacing, CBF decreased to 64 +/- 24% of baseline and increased to 265 +/- 100% of ischemic flow after adenosine administration. In contrast, pacing increased MBF to 184 +/- 66% of baseline, measured as a function of the increased rate-pressure product (r = 0.69; P < 0.05). Thus, during pacing, perfused myocardial mass drastically decreased from 30 +/- 23 to 12 +/- 11 g (P < 0.01). Distal coronary pressure remained stable during pacing but decreased after adenosine administration. Stenting increased perfused myocardial mass to 39 +/- 23 g (P < 0.05 vs. baseline) as a function of the increase in distal coronary pressure (r = 0.71; P < 0.02). In conclusion, the vasoconstrictor response to pacing-induced ischemia is heterogeneously distributed and excludes a tissue fraction from perfusion. Within perfused tissue, the metabolic demand still controls the vasomotor tone.