<p>This study investigates signal diffraction as a significant, yet often unmodeled, error source in high-precision Global Navigation Satellite Systems (GNSS), which typically rely on carrier phase measurements for sub-centimeter accuracy. While standard post-processing mitigates major errors like orbits, clocks and atmospheric delays, localized effects such as diffraction remain a challenge. Diffraction introduces a slowly varying, systematic bias into the measurements, which degrades the accuracy of high-precision positioning. The error is deterministic and structured, meaning it cannot be averaged out and requires physical or empirical modeling. By analyzing double-difference carrier phase residuals, precise point positioning (PPP) position estimates and signal-to-noise ratio (SNR) measurements from both long-term real and synthetically modified datasets, the research demonstrates that diffracted signals are indeed received in areas obstructed from direct satellite line of sight. These bent signals, when inadvertently included in positioning solutions, introduce significant errors that standard algorithms fail to correct, as they are highly dependent on the antenna’s immediate local environment. Our simulation results show that the scatter of the position estimates can get twice as large or even more in the presence of non-minimal diffraction effects. A similar degradation can be seen from the analysis of real data, underscoring that diffraction is a critical factor potentially compromising the utmost accuracy of precise GNSS applications.</p>