This thesis deals with separation of freeway traffic using dynamic lane assignment, based on ones destination; either or not to the next downstream exit. At freeway exits a high amount of lane changes lead to capacity reduction. Furthermore, when the flow to the exit exceeds the exit capacity (like at an IKEA on Sundays) a queue will form that spills back to the freeway. In case no measures are taken, congestion spreads out over all freeway lanes. Traffic flow theory shows that separating exiting traffic from through-going traffic can prevent this total roadway blockage in case increasing the exit capacity is not possible. Compared with existing measures, this study explores chances and drawbacks of dynamic flow separation on one roadway. Here dynamic means variable in time and space, so no physical or static separation is used. The goal of this type of separation is to improve outflows for through-going traffic near oversaturated off-ramps. Therefore two traffic controllers have been designed in this study. Through-going vehicles are guided away from the rightmost lane while exiting vehicles are guided to the rightmost lane. The dynamic aspect is that the length of this separation measure is based on the location of the tail of the queue. Both controllers switch on when the vehicle speeds drop below a threshold value at a specified location. In the first control strategy (a feedback controller) the length of the dynamic separation upstream of the exit is determined by a fixed offset distance upstream of the measured tail. In the second control strategy (a feed forward controller) the location of the tail is predicted using shockwave estimation. The separation length is now determined by the shockwave speed and direction. Both controllers have been tested in an adapted version of FOSIM for different flow/capacity ratios by altering controller parameters like intervention location and (initial) offsets. The simple feedback controller mostly improves the outflow for through-going traffic. The advanced feed forward controller works as well, but the control behavior is very unstable due to measurement errors of many variables. The benefit in outflow can grow to 30% with a well specified intervention location, offsets around 1000 m and compliances from 80%. The results show a more uniform traffic situation near the exit, reduction of congestion spillback and no total roadway congestion after implementing a controller. A clear separation in flows can be seen resulting in higher speeds on the leftmost lane. The results are suboptimal however, because the speeds and flows on the lane adjacent to the exit queue are low. This is due to the legal and safety aspects related to the maximum speed difference between non-physically separated lanes.