A pair of hydrographic sections (35¿N, 155¿W) were analyzed to compute absolute velocity by using a variation of the technique by Stommel and Schott (1977). Absolute velocity is determined from an integrated form of the potential vorticity equation by a technique suggested by Davis (1978). This study is the first application of this technique that allows a direct comparison between the uncertainty in estimating a smooth density field and the amount of imbalance in the system of model equations. Because the amount of incompatibility (imbalance) in the system of equations is far smaller than is allowed by the uncertainty in defining the smooth field, the model equation is considered adequate for this set of data. Below 400 m, the nearly constant zonal isopycnal slope indicates that potential vorticity is uniform on isopycnal surfaces. Since the method depends on resolving flow directions from the intersection of isopycnals and surfaces of constant potential vorticity, the absolute velocity is indeterminate in this region. The model equation does, however, constrain the structure of the meridional density field and requires a poleward shift in the latitude which successively deeper isopycnals reach their maximum depth. The fact that this poleward translation can be predicted over several degrees of latitude suggests potential vorticity is uniform over a substantial portion of the North Pacific subtropical gyre. This poleward translation of the density field is an aspect of subtropical density fields, in general, and occurs in conjunction with a translation in the field of geopotential anomaly. It is directly related to the curvature in the deep portion of the beta spiral.