We develop a model to test the hypothesis that a high fluid pressure pulse and subsequent fluid flow caused a spatially extensive increase in the VP/VS ratio following the Mw = 8.0 Antofagasta, Chile, subduction zone earthquake. The postseismic anomaly appeared within a 50-day period inside the forearc region of the Andes continental crust. We model this anomaly with a poroelastic medium that combines the fluid flow response to mean stress changes from slip along a dislocation plane, with a pore pressure pulse initiated by the coseismic rupturing of a seal separating hydrostatic-lithostatic fluid pressure conditions in the hanging walls and footwalls of the subduction zone, respectively. Variations in seismic velocity due to porosity and pore pressure changes are calculated using Gassmann's formula and the empirical law of critical porosity. We show that the slip-induced perturbation of the mean stress field is insufficient to explain the anomaly, but the release of lithostatic pressurized fluid trapped below the rupture plane (within the oceanic crust) can explain the transient changes in seismic velocities. Our preferred model requires a very large intrinsic permeability of around 1 ¿ 10-13m2, suggesting a mechanism of a high-amplitude pressure pulse propagating through a highly permeable fracture system of the lower continental crust.