Seafloor gravity and pressure measurements for 4D reservoir monitoring require precise models of the time-varying tidal signals. Current seafloor instrumentation can resolve in time-lapse gravity differences and in pressure. To verify model accuracy, a seafloor gravimeter and pressure gauge were operated continuously for next to the Troll A gas platform in the North Sea (60.64227° north, 3.72417° east) at a depth of . The seafloor gravity and pressure time series were filtered and corrected with estimates from the tidal model, which predicts the solid earth tide, ocean loading, and direct gravitational attraction of the varying water level. The rms difference between the observed tidal gravity signal and the prediction is about during periods when there are no surface storms. A slight difference is observed for the direct attraction of the water overhead as computed from the tidal prediction versus that computed from direct seafloor pressure measurements when the entire record is analyzed; it shows an rms difference of , equivalent to of water-height variation, yielding a gravity effect of . We conclude that existing theoretical tide models in combination with in situ pressure records are sufficiently precise for correcting time-lapse gravity observations.