a b s t r a c tConventional 3D seismic data provides direct evidence for glacial influence during the early Pleistocene sedimentation in the Central North Sea. We identify iceberg ploughmarks as dim linear to curve-linear features in three early Pleistocene horizons that have high reflection amplitude compared to adjacent horizons. The anomalous horizons are interpreted to be reflections from thin sandy layers saturated with gas. The gas acts as a contrast liquid illuminating the thin sand layers. The reason for this is the difference in acoustic properties between water and gas saturated sand layers. The combination of thin bed effects and shallow gas makes the iceberg ploughmarks easily detectable as dim features in seismic reflection amplitude maps. Our interpretation is based on analysis of real seismic data, well logs and modeled seismic response. The methods we use include interpretation of horizons followed by extraction of reflection amplitudes, well log analysis, pre-stack amplitude versus offset analysis of high resolution 2D seismic data and time-lapse analysis of seismic. Seismic modeling is performed to study interactions between thin sand beds, shallow gas and iceberg ploughmarks. A new trapping mechanism for shallow gas is presented and seismic modeling of this trap strengthens our interpretation. The trap is created by iceberg ploughmarks in sandy layers that are covered by finer and less permeable sediments. For this area we find that conventional seismic interpretation is superior to the much used method of studying seismic time-slices for detection of iceberg ploughmarks, both with respect to time and detectability. This study shows that the interpreter should look for high amplitude horizons with amplitude variations laterally when trying to detect iceberg ploughmarks.