Bubble screens are used at sea locks to mitigate salt intrusion into inland water systems. In this paper the effectiveness of a bubble screen in delaying the mixing of salt and freshwater via lock exchange was studied. Laboratory-scale experiments investigating the flow field and mixing caused by a bubble screen are presented. The tests include both the homogeneous situation of freshwater on both sides of the screen and the inhomogeneous situation where there is an initial density difference across the screen, which leads to a density current after the lock gate is removed or opened. Optical measurement techniques were applied, giving spatially detailed flow velocities and densities. The parameters varied between tests are the airflow discharge and the bubble size. The results show that the bubble size in the screen had a significant effect with a screen with bubbles of 1-2 mm being more effective at generating a surface flow in the homogeneous case but less effective at keeping the fresh and salt sides separated in the inhomogeneous case, when compared with a screen of 4-6 mm bubbles. The point of maximum effectiveness for separating salt and fresh sides was also shown to be dependent on bubble size.