SETTING: England and Wales. OBJECTIVE: To quantify the relative contribution of vaccination, chemotherapy and preventive therapy to the reduction in tuberculosis incidence in England and Wales between 1953 and 1990. DESIGN: A compartmental model of tuberculosis transmission was fitted to notification data between 1913 and 1939 to estimate pre-vaccination parameters. Best-fit estimates of the rates of chemotherapy and preventive therapy were derived by fitting the model to notification data between 1953 and 1990. Published vaccination rates were used. MAIN OUTCOME MEASURE: Number of cases of pulmonary tuberculosis averted. RESULTS: The numbers of respiratory tuberculosis cases averted between 1953 and 1990 by the use of preventive therapy, vaccination and chemotherapy were 288318, 57085 and 206996, respectively. CONCLUSIONS: Of those interventions considered, preventive therapy has the greatest impact on transmission. The duration of infectiousness is long, with an onset that is likely to pre-date sputum positivity.