The Geodesy Advancing Geosciences and EarthScope (GAGE) Facility Global Positioning System (GPS) Data Analysis Centers produce position time series, velocities, and other parameters for approximately 2000 continuously operating GPS receivers spanning a quadrant of Earth’s surface encompassing the high Arctic, North America, and Caribbean. The purpose of this review is to document the methodology for generating station positions and their evolution over time and to describe the requisite trade-offs involved with combination of results. GAGE GPS analysis involves formal merging within a Kalman filter of two independent, loosely constrained solutions: one is based on precise point positioning produced with the GIPSY/OASIS software at Central Washington University and the other is a network solution based on phase and range double-differencing produced with the GAMIT software at New Mexico Institute of Mining and Technology. The primary products generated are the position time series that show motions relative to a North America reference frame and secular motions of the stations represented in the velocity field. The position time series themselves contain a multitude of signals in addition to the secular motions. Coseismic and postseismic signals, seasonal signals from hydrology, and transient events, some understood and others not yet fully explained, are all evident in the time series and ready for further analysis and interpretation. We explore the impact of analysis assumptions on the reference frame realization and on the final solutions, and we compare within the GAGE solutions and with others.