In this study, we quantified systematic errors in surface gravity anomalies, which were caused by systematic
errors in gravity and heights of the gravity stations, and computed their impact on the quasi-geoid model of the
Netherlands and Belgium. We found that 70% of the gravity datasets have statistically significant biases ranging
from −2 mGal to 1.5 mGal. The primary impact of the biases are long-wavelength systematic distortions in the
quasi-geoid model with a peak-to-peak amplitude of 8 cm. We also found systematic errors in the height networks
of the Netherlands and Belgium, which cause corresponding errors in the heights of the gravity stations. They range from −3.0 cm to 1.7 cm and −12.0 cm to 5.0 cm, respectively. They also introduce errors in the transformation parameters to EVRF2007 of several centimetres. However, the impact of the height errors on the quasi-geoid model is negligible with a peak-to-peak amplitude of less than 0.1 cm.