Mascon products derived from Gravity Recovery and Climate Experiment satellite gravimetry data are widely used to study the Greenland ice sheet mass balance. However, the products released by different research groups—JPL, CSR, and GSFC—show noticeable discrepancies. To understand them, we compare those mascon products with mascon solutions computed in-house using a varying regularization parameter. We show that the observed discrepancies are likely dominated by differences in the applied regularization. Furthermore, we present a numerical study aimed at an in-depth analysis of regularization-driven biases in the solutions. We demonstrate the ability of our simulations to reproduce 60%–80% of biases observed in real data, which proves that our simulations are sufficiently realistic. After that, we demonstrate that the quality of mascon-based estimates can be increased by a proper modification of the applied regularization: no correlation between mascons is assumed when they belong to different drainage systems. Using both simulations and real data analysis, we show that the improved regularization mitigates signal leakage between drainage systems by 11%–56%. Finally, we validate various mascon solutions over the SW drainage system, using trends from (i) the GOCO-06S model and (ii) the Input-Output Method as control data. In general, the in-house computed trend estimates are consistent with the trends from CSR and JPL solutions and the trends from the control data.