The thesis explored here is the use of pth data for comparing primitive equation models with observations. Such comparisons naturally focus on the inversion of path data for fundamental kinematic characteristics and subsequent descriptive interpretations of both models and observations. The analysis here focuses on uncertainties that may affect this interpretation. These include random position errors and numerical noise introduced by the nonlinear inversion ansatz as well as dynamical processes that introduce both low- and high-frequency variability in the kinematic fields. After each of these effects is assessed, the inversion ansatz is applied to paths generated from an eddy-resolving general circulation model of the Gulf of Mexico. Model Loop Current ring centers determined from different paths are found to be consistent and in good agreement with observations in the abyssal gulf. In addition, these centers are compared with the model surface and interface pressure anomaly centers. The simulated paths showed ejection of mass from the Loop Current ring as it interacted with the western slope. Observed and simulated ring kinematics inferred from appropriate paths are also made and compared with observations for the slope region. The paper concludes with a discussion of the potential for using simulated and observed path data for extracting flow physics and quantifying comparisons between models and observations. ¿ American Geophysical Union 1990