After coarctectomy, local loss of distensibility is noted in addition to mild anatomic narrowing. We hypothesize that the increased Doppler peak velocities measured at the aortic isthmus in these patients partly reflect obstruction secondary to the stiff surgical scar. The hypothesis was studied in a pulsatile hydraulic model. Thirty-one patients (13.0 +/- 4.0 years of age), 10.5 +/- 4.7 years after coarctectomy by end-to-end anastomosis, were studied clinically and echocardiographically. Indexes of distensibility were calculated. The effect of isolated increased stiffness was studied in vitro with a stiff and a compliant 1:1 scale latex model of the aorta mounted in a pulsatile full-scale circulation loop. Local stiffening was obtained by a rigid ring mounted around the aorta, fitted to the dimension of the unloaded aorta. For different pressure and flow regimens, pressures and Doppler velocities were measured across the ring. Mean peak velocities at the surgical scar were 2.2 +/- 0.4 m/s. Mild anatomic stenosis was present. All distensibility indexes indicated locally increased stiffness (P <.001). In the stiff latex model, Doppler peak velocities increased from 1.89 +/- 0.04 m/s to 2.32 +/- 0.06 m/s (P <.03); in the compliant model, from 1.15 +/- 0.03 m/s to 1.79 +/- 0.05 m/s (P <.001). The increase of Doppler peak velocities depends on model compliance only and is independent of flow rate, length of the noncompliant segment, and viscosity of the perfusion fluid. Velocities do not change when semicircular stiffening is applied. We have demonstrated in vitro that isolated local nondistensibility leads to vessel narrowing during vascular distension. The relative contribution of local scar stiffness in the increase of Doppler peak velocities after coarctectomy was hereby assessed.