Bordetella pertussis is an exclusively human pathogen found worldwide in all populations. Complications of pertussis including bronchopneumonia, failure to thrive from post-tussive vomiting, cerebral hypoxia leading to brain damage, and death are strongly concentrated in infants. The similarity in mortality profile with age between countries that are markedly different with respect to demography, health systems, socioeconomic status, and surveillance systems is striking. It is because infants are most likely to suffer serious complications of infection or die that prevention of pertussis in infants is the primary goal of immunization programmes. Of all the vaccine-preventable infections, pertussis remains the most epidemiologically challenging to understand. This is partly because of surveillance issues but also the lack of an established correlate of protection that allows susceptible and immune individuals to be distinguished. These factors are compounded by our imperfect understanding of the mechanism of protection from acellular and whole-cell vaccines, and the product-specific differences in efficacy and effectiveness. It can therefore be difficult for policymakers to be confident about the optimal number and timing of primary doses and how many and at what age booster doses should be given in their setting, as reflected in the plethora of schedules in use throughout the world. There are even greater challenges for the World Health Organization when attempting to make global policy as lessons learned in one setting may not appear to apply in others. What is clear is that high coverage early in infancy with three doses of an effective vaccine greatly reduces pertussis mortality and severe morbidity in all settings and that many infants in resource-poor settings are still deprived of that benefit.