Disordered traffic flow includes directed vehicular and non-vehicular flows on surfaces with a continuous lateral dimension such as mixed traffic flow in developing countries, bicycle flow on broader bikepaths, and athlete flow in running, cycling, inline-skating, or cross-country ski marathons. As in pedestrian flow, the motion is fully two-dimensional. However, as in regular vehicle flow, the agents are fast and collisions are harmful and should be avoided. When only bottlenecks, congestions, and travel times are of interest, two-dimensional macroscopic models can be simplified to an effectively one-dimensional Lighthill-Whitham-Richards model with a space and time dependent fundamental diagram. When interactions and safety aspects play a role, microscopic agent-based models are the model class of choice. They combine the force-based philosophy of the Social-Force Model for pedestrians with a car-following model to which these models revert in the spacial case of single-file flows. When adding a suitable floor field, these models revert to a fully two-dimensional integrated car-following and lane-changing model with continuous lane changes. As an example, we present the Intelligent Agent Model (IAM) and simulate it in lane-based city scenarios and lane-free bicycle traffic flow.