Already since the start of civilisation dredging is carried out to create and maintain waterways and ports or to create new land. In recent times the scale of dredging and awareness of potential environmental impact of dredging have increased drastically. An often used dredging vessel is the trailing suction hopper dredger (TSHD). A TSHD pumps up a sediment-water mixture from the bed into a hopper. In this hopper the sediment is given time to settle and the process water is spilled overboard, often through a vertical shaft called the overflow. The spilled process water will contain some suspended sediment which has not deposited yet and this forms a turbid plume. Increased turbidity and deposition at the bed of the suspended sediment from the overflow dredging plume can have negative environmental impact. Identify the factors which create the surface plume and testing a solution to improve this is done in this thesis. Initial mixing of the overflow dredging plume under/near the TSHD is not well understood. Although the plume starts under the keel of a TSHD, it has initial downward velocity and it is denser than the ambient water, sometimes a part of the plume flows up- ward and reaches the free surface right behind the TSHD. This so called surface plume can stay suspended for long periods and is therefore important for the potential environmental impact. This thesis reports on laboratory experiments in still ambient water and numerical modelling with a crossflow. Specific attention is given to the change of overflow shape, from round to rectangular, which is tested in the laboratory experiments. Firstly, all influence factors which create a surface plume are investigated to see which factors have the highest influence. Based on the wish of a passive solution and the ability to do laboratory experiments, the change of overflow shape is chosen to investigate further. The analytical background of plume dispersion with different shapes is investigated to find the differences between them. The laboratory experiments where conducted at the dredging lab at the Technical University of Delft. In this lab, a 25m³ tank filled with water where at the top a frame is positioned which holds the overflow pipe. By adding two type of nozzles (different rectangular aspect ratio) onto the pipe, the outflow shape is changed to see if a rectangular outflow shape improves the plume path in still ambient water. This is done by measuring the concentrations at the bottom and measuring the angle of the plume. Results of the experiments show less entrainment and higher concentrations in the middle of the plume when having a rectangular outflow shape. To say something about the improvement of the plume path in real live, the plume is modelled by adding a crossflow. The information from the experiments is used to create a plume trajectory of the round outflow shape and rectangular outflow shape which are then compared with each other. Practical values are then scaled and used in the model. Results of the model show similarity from the experiments, with a deeper plume trajectory when having a rectangular outflow shape.