This study uses data from a recent survey of New Zealanders' travel behaviour to estimate the pedestrian risk of road accident injury. The estimated risks of accident disaggregated by sex and age were examined by combining road accident data with survey data using the exposure measures "time spent walking" and "number of roads crossed". The two resulting measures of risk were compared both with one another and with the most common mode of presenting of pedestrian accident statistics, accidents per capita. The overrepresentation of the elderly in pedestrian accident statistics was examined in light of their greater susceptibility to injury from a given accident. An attempt was made to adjust for this susceptibility using estimated numbers of "severe impacts" based on the risk of road accident fatality per road crossed. The relative importance of walking as a mode of transport was examined using the travel survey data. Finally, the risks of road accident when crossing at a zebra (unsignalized) crossing were compared with the risks of crossing elsewhere.