Given the apparent greater use of health care services by smokers and predictions of higher costs for a never-smoking population, we aimed in this study to determine whether the acute hospital costs in the last years of life of never- and ever-smokers differed before death using a database of the decedents in 1 year in Hong Kong.To compare the acute hospital use of ever- and never-smoker decedents.The data on cause of death, personal characteristics, and public hospital discharges were linked for all decedents in 1998 in Hong Kong. The incidence rate ratio was used to compare, for ever- and never-smokers, the number of days spent in an acute hospital over the prior 7 years. Analyses were done using specific and all cause mortality, adjusted for sex, lifestyle factors, and life expectancy.Compared with never-smokers who died of the same condition, ever-smokers who died of chronic obstructive pulmonary disease used 28% more acute hospital bed days and those dying of smoking-related cancer 9% fewer. These differences cancelled out over the case-mix of deaths with no net difference in acute bed day use by smoking status for all-cause mortality.There was no difference in acute hospital bed days in the last years of life of ever- and never-smokers but some differences by cause of death. Reducing smoking in this population will not increase acute hospital use.