Direct comparison of myocardial perfusion tracers has been made difficult by variability in experimental models, and by a virtual absence of data comparing tracer uptake to myocardial blood flow under conditions of increased myocardial oxygen consumption, similar to what occurs with dynamic exercise.