The cost-effectiveness of preventive yellow fever vaccination versus emergency mass vaccination campaigns for epidemic control remains a matter of controversy. Until recently, Nigeria and other anglophone countries in West Africa most severely afflicted by yellow fever epidemics have followed a policy of emergency control. The effects of including yellow fever 17D vaccine in the Expanded Program of Immunization (EPI) on the immune status of the Nigerian population was studied under conservative assumptions of vaccine coverage and efficacy. The model defined the age-specific prevalence of immunity resulting from vaccination of infants and from natural endemic infection beginning in 1991 and extending over a time horizon of 35 years. The data were used to predict the number of cases and deaths during hypothetical epidemics in 2006 and 2026, representing the historic periodicity of epidemics. A second model was used to demonstrate that a > or = 60% prevalence of immunity would preclude epidemic yellow fever transmission; under base case assumptions, this prevalence would be reached after 18 years of initiating routine yellow fever vaccination in the Guinea savannah zone, the region most often affected by epidemics. Using assumptions based on data from other African countries, the cost of adding yellow fever vaccine to the existing EPI was estimated as +0.65 per fully immunized child, whereas the cost of emergency vaccination in the face of an epidemic was estimated as +7.84/person. Vaccine coverage rates achievable by the EPI were modeled on recent successes with measles vaccine, and began in 1991 at 60%. The effective vaccine coverage rate in an emergency campaign was taken as 10%, based on recent experience. For an epidemic of moderate size in 2006 (morbidity similar to the documented outbreak in 1987), the cost-effectiveness of emergency mass immunization for control of hypothetical yellow fever epidemics was two-fold higher ($381/case and $1,904/death prevented) than that of the EPI ($763/case and $3,817/death prevented). However, despite its higher cost, the efficiency of the EPI was seven-fold greater in terms of cases and deaths prevented. In large epidemics, such as that occurring over successive years (1986-1991) in Nigeria, cost-effectiveness of the EPI exceeded that of emergency control. The EPI may also play an important role in the prevention of endemic yellow fever. Assuming annual rates of endemic yellow fever predicted by serologic surveys, routine vaccination would significantly reduce morbidity and mortality at cost-effectiveness ratios within the range for other diseases prevented by the EPI, including polio, tetanus, and diphtheria.(ABSTRACT TRUNCATED AT 400 WORDS)