Background-Failing hearts display slow relaxation with apparent increased load sensitivity. However, inaccuracies of monoexponential analysis can contribute to these observations, and different qualitative and quantitative results might be obtained by alternative models. We tested whether pressure relaxation of failing hearts consistently deviates from a monoexponential waveform, leading to overestimations of lusitropic change and load sensitivity by monoexponentialderived time constants. Methods and Results-Fourteen dogs were studied before and after tachycardia pacing-induced heart failure. Relaxation time constants were derived by monoexponential fits ( E ) with zero or nonzero asymptotes and by a logistic fit ( L ). L assumes nonlinear relations between pressure and its first derivative, whereas E assumes a linear dependence. Load sensitivity of was tested by comparing beats during vena caval occlusion. E prolonged by 75% to 80% with heart failure, 3 times more than L (PϽ0.01). E displayed marked load sensitivity in failing hearts, shortening during preload reduction, whereas L was little changed by the same loading maneuver. Neither L nor E varied with preload in control hearts. The discrepancy between E and L results was due to nonmonoexponential decay reflected by nonlinear pressure-time derivative of pressure plots, which was enhanced with heart failure (PϽ0.01). This nonlinearity was reduced by -adrenergic stimulation, lowering preload sensitivity of E to control levels. Conclusions-Isovolumic relaxation in failing hearts deviates from a monoexponential waveform, leading to overestimated relaxation delay and increased load sensitivity of monoexponential time constants. This deviation is under -adrenergic modulation. The logistic model improves the fit-to-real pressure decay in failing hearts, providing more stable measures of relaxation. (Circ Heart Fail. 2010;3:268-276.)