In this contribution, we analyze the impact of different GPS processing strategies on ocean tide loading estimation. We use continuous GPS data acquired during a 4-month campaign performed in 2004 in Brittany, Northwest France. Since the expected geodynamical signal in the estimated positions is exceeding the typical GPS data analysis noise, this data set can be used to compare the results obtained with different analysis software packages. Moreover, in this specific case we need short sub-daily solutions to study short-period signals instead of classical 24 h-solutions. The GPS capability for measuring 3D ocean tide loading deformation has already been assessed, but since we are looking for the finest signal as the one induced by the shallow water constituents, it is essential to be sure that the position time series represent a geodynamical signal and are not biased by the data processing strategy used. To analyze the possible effect of the methodology used on the geodynamical results, we compare different solutions computed with different strategies (Double Differencing and Precise Point Positioning) with various GPS analysis software packages (Bernese, GAMIT, GINS, and GIPSY/OASIS). We show that the different solution consistency is at the level of 1–3 mm. We also show that the data processing strategy has a mean effect of about 10–20% of the ocean tide loading signal amplitude.