The fault restoration technique of De Chabalier and Avouac <1994> is applied to an ultra-high-resolution bathymetry data set from the East Pacific Rise (EPR) at 18¿14'S. Fault offsets are calculated and subtracted from the original seafloor bathymetry to examine the morphology of the unfaulted seafloor surface within an area encompassing the ridge axis 400 ¿ 1600 m in dimension. The restored topography reveals a gently sloping seafloor ~200 m wide, which slopes 5¿ inward toward the spreading axis. We attribute this inward sloping seafloor to subsidence within the axial trough due to subsurface magmatic deflation. The vertical deformation field extracted from the bathymetry is used to characterize the ridge axis fault population present in the area. Median fault throws (9 m for inward-facing and 8 m for outward-facing faults) are comparable to values measured for nearby mature abyssal hill fault populations, suggesting a genetic link. However, median fault spacings (70 and 46 m) are an order of magnitude smaller, and estimated total extensional strain is 3¿--4¿ greater than values measured for ridge flank faults. These differences indicate the axial fault population at 18¿14'S cannot be rafted onto the ridge flanks to form abyssal hill faults without significant modification, presumably via volcanic burial. We attribute the dense faulting observed in this area to slip on axial fissures during subsidence of the crestal region. The surface subsidence of a volcanic edifice can be modeled in terms of volume change in the magma source reservoir and volume of magma withdrawn from the reservoir. Using the relationship derived for a sill-like magma body, we estimate that the axial depression at 18¿14'S could represent magma withdrawal associated with a small number (4--22) of the frequent dike injection and eruption events required to build the upper crust during the evolution of the trough. The subsidence volumes represented by the axial topography at 18¿14'S are a small percentage of the underlying upper crustal sections (3--4%), suggesting that only a minor mismatch between magma recharge and withdrawal from the axial melt lens during ongoing plate separation could account for this pronounced axial depression.