"Critical coronary stenosis" (reduction of coronary blood flow [Q] until reactive hyperaemia following 15s coronary occlusion is just abolished) and "functional stenosis" (reduction of Q until systolic shortening (dL) of post-stenotic myocardium is curtailed by 50%) were compared with respect to the reduction in Q necessary and the effect on regional myocardial and global ventricular function. In 9 anaesthetized (piritramide) dogs, enddiastolic length (edL) and dL of a myocardial area supplied by the left descending coronary artery (LAD) were measured by sonomicrometry. Left-ventricular end-diastolic pressure (LVEDP) and dP/dt, aortic pressure (AoP), stroke volume (SV), and heart rate (HR) were monitored. QLAD was stepwise reduced by a snare. Critical stenosis, present at a 25% reduction of QLAD, had no effect on regional and global ventricular function, but recovery of dL after release of 15s LAD occlusion was significantly delayed. Further obstruction of QLAD progressively impaired dL, reaching 50% dL (functional stenosis) by a flow reduction of about 50%. The decrease in dL was accompanied by an increase in edL. The haemodynamic effects of the functional stenosis were rather discrete (LVEDP + 5%, SV-12%, dP/dtmax-8%). Models of myocardial ischaemia used to study the effect of drugs or other haemodynamically effective interventions should be able to show functional impairment as well as improvement of the ischaemic myocardium. The critical stenosis does not impair myocardial function and, consequently, a favourable influence on the function and, consequently, a favourable influence on the function of ischaemic myocardium by any intervention may not become evident. However, in the presence of a functional stenosis the change in systolic shortening of the ischaemic myocardium is a very sensitive response to any intervention which affects the energy balance of the ischaemic myocardium. Therefore, this model should be preferred.