Aims: To improve estimates of deaths from pertussis in England and to identify reasons for under ascertainment. Methods: Comparison of deaths identified from enhanced laboratory pertussis surveillance, hospital episode statistics (HES), and Office for National Statistics (ONS) death registrations in England, 1994 to 1999. Results: A total of 33 deaths were identified, 88% of infants less than 4 months old. There was overlap between all sources; 22 deaths were identified in the enhanced pertussis surveillance system, 18 in ONS mortality data, and nine in HES. Children who had died from pertussis without mention of pertussis on the death certificate were more likely to have been certified by coroners than those with mention of pertussis (p = 0.0005). Using capture-recapture analysis, the total number of deaths from pertussis in the five and a half year period is estimated to be 46 (95% CI 37 to 71), or around nine deaths per year. Conclusions: National mortality statistics significantly under estimate deaths from pertussis in England and are inadequate for monitoring the national immunisation programme. The largest number of deaths is identified through enhanced laboratory surveillance. Death registration systems should take into account available microbiological information to ensure that cause of death is accurately assigned.