Phase Space Density (PSD) matching can be used to identify the presence of nonadiabatic processes, evaluate accuracy of magnetic field models, or to cross-calibrate instruments. Calculating PSD in adiabatic invariant coordinates requires a global specification of the magnetic field. For a well specified global magnetic field, nonadiabatic processes or inadequate cross calibration will give a poor PSD match. We have calculated PSD(μ, K) for both Van Allen Probes using a range of models and compare these PSDs at conjunctions in L* (for given μ, K). We quantitatively assess the relative goodness of each model for radiation belt applications. We also quantify the uncertainty in the model magnetic field magnitude and the related uncertainties in PSD, which has applications for modeling and particle data without concurrent magnetic field measurements. Using this technique, we show the error in PSD for an energy spectrum observed by the relativistic electron-proton telescope (REPT) is a factor of ∼1.2 using the TS04 model.