Structural health monitoring (SHM) is becoming increasingly important due to the large network of aging infrastructure in the United States. Monitoring of such structures is traditionally accomplished using accelerometers or strain gages, which are unable to resolve the global displacements of the structure, including static, quasi-static and dynamic components. Given the need to observe these various response components, global positioning systems (GPS) have become a viable alternative for SHM. While GPS has been successfully applied in full-scale deployments, several practical considerations must be assessed, including: nonstationarity of the GPS reference site, the use of multi-reference processing, potential accuracy loss in real time kinematic (RTK) mode and the effects of multipath errors. The final issue is of most concern, as it poses the largest untreatable error source for GPS monitoring in dense urban zones. Full-scale data taken from a current GPS deployment in Chicago is provided to show the promise of GPS, while illustrating the distortions induced by multipath. Thus, a calibration program is developed in this study to address the aforementioned practical considerations. A comprehensive assessment of the performance of a GPS sensor versus two traditional sensors' an accelerometer and a terrestrial positioning system (TPS)' is provided. Finally, the multipath effects are quantified and removed from the post-processed GPS measurements through a controlled testing program and the use of time-series detrending to generate a GPS Distortion Signature to eliminate both multipath effects and other long-period distortions from changing satellite orientations. The end result is a better understanding of GPS technology and practical tools that will enable its effective use for SHM, even for deployments in dense urban areas.