The ability to understand and predict the potential epidemiological impact of a vaccination program has been greatly enhanced by the use of mathematical models of disease transmission. Vaccines have the potential to produce major changes in disease epidemiology resulting both from direct protective effects and indirect herd immunity effects. The use of models to understand the population biology of organisms-such as the meningococci and pneumococci-has the potential to undergo serogroup-type switching, and the relationship between carriage and disease-causing isolates is likely to be a major application in the future with the arrival of new polysaccharide conjugate vaccines. The choice of an immunization strategy will be highly dependent on the biological properties of the vaccine, together with the epidemiological and biological characteristics of the natural infection. Underpinning all models requires sound baseline data on the epidemiology of disease, age-specific disease incidence, and antibody prevalence profile in a population prior to embarking on any immunization program.