Plastic pollution of our oceans is becoming a bigger problem every day. The production of new plastics is extremely cheap and convenient resulting in a lack of incentive to retrieve plastic litter. Annually 8 billion kg of plastics enter the ocean polluting the ocean environment, of which the largest part via rivers and other waterways. Eventually, plastic enters the food chain through fish and other sea life. Organisations are arising worldwide to reduce the plastic stream to the ocean. The Great Bubble Barrier is based on the idea of filtering plastics from rivers by making use of a bubble curtain. The principle behind this idea is as follows: a long tube is placed diagonally across a river and air is pumped through it creating a wall of bubbles. The rising bubbles enhance upward entrainment of the surrounding water. When approaching the surface the vertical flow is transferred into horizontal flow away from the bubble curtain on both sides. Together with the crossflow, this horizontal flow induced by the bubbles drives the plastic to the side. This horizontal flow is strongest at the surface and decays linearly down to 25% of the water depth. Therefore, only the plastic litter floating in the upper layer is affected. The vertical and horizontal flow induced by the bubble curtain in still water has been extensively researched in the last century. In this thesis, the influence of the crossflow on the behaviour of the bubble curtain is explored. Predictions of the horizontal surface flow are based on a theoretical model and validated with experiments performed at the Eastern Scheldt Flume at Deltares. The model is based on the assumption that the maximum bubble induced horizontal velocity is not influenced by the strength of the crossflow. At every depth in the surface layer, the two velocities can be summed up. Depending on their magnitude and the angle of the curtain, the direction of the resulting flow at that depth can be calculated. If the resulting flow is directed to the upstream side of the bubble barrier, the plastics are assumed to stay on the upstream side too and will, therefore, be led to the side of the river. If the resultant is directed to the downstream side of the bubble curtain the plastics are assumed to break through. The vertical velocity of water in the plume, as well as the horizontal bubble induced velocity, is calculated using two equations. One is derived from the momentum balance and the second is empirically found by Bulson. The results of both equations are compared with the experiments. For the vertical velocity, the momentum balance derived equation gives better results, whereas Bulson's equation gives a higher accuracy for the horizontal velocity. The crossflow is described by a logarithmic profile. For all angles and crossflow velocities, the airflow required to keep the resultant upstream can be calculated at every depth in the surface layer. The airflow that is theoretically required to keep the resultant at a 10 cm depth upstream, is shown to be sufficient to block 90% of the tested objects in all experimental setups. The smaller the angle relative to the direction of the crossflow, the smaller the total required airflow over the whole tube despite the longer required tube length. This work has gained a better understanding of the working principle of a bubble curtain as a plastic barrier. Further research on its performance in deeper waters is advised.