To help the Children's Vaccine Initiative (CVI) achieve its goal of new and improved children's vaccines, we developed and applied a cost-effectiveness model to set priorities for vaccine development. The model measures the health benefits in additional Quality-Adjusted Life Years (QALYs) gained by the combined birth cohorts of all developing countries over an assumed useful life of a proposed vaccine (generally 10 years). It measures costs as the net cost of developing, procuring, and administering the vaccine to the same population and time frame compared to the status quo (the current vaccine, if any). It weights each dollar of in-kind allocation of the existing health infrastructure less heavily than a dollar cash outlay to purchase new vaccine to reflect severe constraints on foreign exchange and non-personnel costs. It expresses cost-effectiveness as the net cost per QALY. The model was applied to 13 candidate vaccines selected by the CVI for initial analysis on the basis of their near-term feasibility. The five most cost-effective improvements, each of which could generate a QALY inexpensively (below $25 per QALY), were an early-administration or an early two-dose measles vaccine, slow release tetanus toxoid (for women), improved typhoid vaccine, and hepatitis B combined with diphtheria-tetanus-pertussis vaccine.