Since construction of ventricular function curves and computation of the maximum velocity of the contractile element (VCEmax) are technically difficult, a number of other empiric contractility indices were compared in (1) chronically instrumented awake dogs and (2) in anesthetized dogs given slug doses of isoproterenol or volume loaded before and after autonomic blockade. Five indices: VCEmax, max dp/dt, t-max dp/dt, (max dp/dt)/LVPd, and (max dp/dt)/ IIT, were tested in unanesthetized animals and were found to correlate well with each other, the coefficient of correlation varying from 0.56 to 0.89. Ten indices: max dp/dt, t-max dp/dt, (max dp/dt)/LVPd, (max dp/dt)/LVPt, peak (dp/dt/LVPt), t-peak (dp/dt/ LVP ), t-max flow, PEP, and VCEmax (left ventricular end diastolic volume [LVEDV] both measured and assumed), were compared in the anesthetized animals. They were found to be satisfactorily reproducible, to respond well to small doses of isoproterenol but to be moderately sensitive to volume loading in the case of PEP, t-peak (dp/dt/LVPt), (max dp/dt)/LVPt, peak (dp/dt/LVPt), max dp/dt and VCEmax (LVEDV assumed None of these indices can be recommended without reservation and none is without merit. PEP can be calculated from data obtained by noninvasive methods and will be useful in the absence of significant mitral or aortic valve disease. VCEmax offers advantages only if ventricular end diastolic volume is measured.