In May 2000, the code governing the conduct of drivers at marked zebra crossings in Sweden became stricter to improve safety and mobility for pedestrians. Accident analysis based on a macro study of Sweden suggests that the injury risk in marked, not reconstructed zebra crossings increased by 27% for pedestrians and 19% for bicyclists. This means that reconstructions are needed for the effect of the change of code to be positive. An information campaign combined with physical countermeasures to encourage drivers to stop at least 5 m before marked zebra crossings is also suggested as well as a code change forbidding unsignalized marked zebra crossings where overtaking is possible, for example on streets with four lanes.