Estimation of the maximum velocity (Vmax) of the contractile element of the intact left ventricular wall muscle demands extrapolation of the force-velocity curve to zero load. The present paper describes our on-line computation system for measuring and analysing pressure derived Vmax using both linear (Vmax-lin) and exponential (Vmax-exp) extrapolation methods. The developed pressure during isovolumetric phase of systole was used as an equivalent of the force. Testing on anaesthetized artificially ventilated dogs showed the exponential function to fit pressure-velocity data better than the straight line did. The Vmax-exp attained 15-35% greater values than Vmax-lin, but both responded almost equally when considered on the basis of linear regression analysis (r = 0.991, n = 725). Changes of contractility caused by i.v. infusion of isoproterenol, calcium chloride or propranolol were practically similar when assessed by either method of Vmax computation, or by dP/dtmax. Volume loading by dextran infusion increased not only dP/dtmax, by 33 +/- 13%, but also Vmax, up to 24 +/-. When arterial pressure was raised by phenylephrine infusion, or heart rate by atrial pacing, dP/dtmax increased significantly while Vmax remained unaltered. Hence, the linear and exponential dP/dtmax increased significantly while Vmax remained unaltered. Hence, the linear and exponential extrapolation procedures provided comparable values for Vmax, but the linear one due to its simplicity is more suited for on-line computation. The Vmax thus obtained is, however, not independent of the changes in preload.