A merge is a point of a highway where two or more streams of traffic flow into one. It is always easy to solve the demand–supply problem at a merge when the merge is operating under uncongested condition. However, when congestion backs up exceeding the merging point where multiple streams of traffic meet, one is typically faced with splitting downstream supply among the merging branches. Solutions of this situation are multiple and several merge queuing models have been proposed in the past. To address the drawbacks of the past models, this paper proposes a capacity-based weighted fair queuing (CBWFQ) model that is characterized by its fidelity (approximation to real situation), simplicity, and extensibility. Based on the CBWFQ merge queuing model, a simplified kinematic wave model is formulated to model traffic operation at a merge bottleneck.