OBJECTIVES: To compare the distances travelled for inpatient treatment in England between electoral wards prior to the introduction of a policy to extend patient choice and to consider the impact of patients' socio-economic status. METHODS: Using Hospital Episode Statistics for 2003-04, the distance from a patient's residence to a National Health Service hospital was calculated for each admission. Distances were summed to electoral ward level to give the distribution of distances travelled at ward level. These were analysed to show the distance travelled for different admission types, ages of patient, rural/urban location, and the socioeconomic deprivation of the population of the ward. RESULTS: There is considerable variation in the distances travelled for hospital admission between electoral wards. Some of this is explained by geographical location: individuals living in more rural areas travel further for elective (median 27.2 versus 15.0 km), emergency (25.3 versus 13.9 km) and maternity (25.0 versus 13.9 km) admissions. But individuals located in highly deprived wards travel less far, and this shorter distance is not explained simply by the closer location of hospitals to these wards. CONCLUSIONS: Before the introduction of more patient choice, there were considerable differences between individuals in the distances they travelled for hospital care. An increase in patient choice may disproportionately benefit people from less deprived areas.