BACKGROUND: Pneumococcal conjugate vaccine (PCV) has been in routine use in the United States for 5 years. Prior U.S. cost-effectiveness analyses have not taken into account the effect of the vaccine on nonvaccinated persons. METHODS: We revised a previously published model to simulate the effects of PCV on children vaccinated between 2000 and 2004, and to incorporate the effect of the vaccine in reducing invasive pneumococcal disease (IPD) in nonvaccinated persons during those years. Data from the Active Bacterial Core Surveillance of the Centers for Disease Control and Prevention (2000-2004) were used to estimate changes in the burden of IPD in nonvaccinated adults since the introduction of PCV (compared with the baseline years 1997-1999). Results combined the simulated effects of the vaccine on the vaccinated and nonvaccinated populations. RESULTS: Before incorporating herd effects in the model, the PCV was estimated to have averted 38,000 cases of IPD during its first 5 years of use at a cost of dollar 112,000 per life-year saved. After incorporating the reductions in IPD for nonvaccinated individuals, the vaccine averted 109,000 cases of IPD at a cost of dollar 7500 per life-year saved. When the herd effect was assumed to be half that of the base case, the cost per life-year saved was dollar 18,000. CONCLUSIONS: IPD herd effects in the nonvaccinated population substantially reduce the cost, and substantially improve the cost-effectiveness, of PCV. The cost-effectiveness of PCV in actual use has been more favorable than predicted by estimates created before the vaccine was licensed.