The paper considers the operation of Zebra and Pelican crossings in the United Kingdom and the associated pedestrian and vehicle delays. The derivation of empirical relationships for saturation flow based on site observations at pedestrian crossings is described and it is shown that saturation flow depends on pedestrian flow and is substantially lower than at signal controlled junctions. Simulation programs are used to calculate vehicle capacity over a range of pedestrian flow for typical Zebra and Pelican crossings.