SummaryResults from the first phase III efficacy trial of an HIV vaccine will be available within the next 2-3 years. Thus, it is imperative to start planning now to address how any effective vaccines should be used. In the absence of definitive information on the characteristics of the first generation of HIV vaccines, the following assumptions were made: the vaccine will (i) have only low to moderate efficacy (on the order of 50%); (ii) not be inexpensive (on the order of 10 to 30 US $ per dose); (iii) require multiple doses; and, (iv) at least initially, be available in limited quantities. A vaccine with that profile would not be suitable for general use in all countries, and it might have to be initially targeted to populations at higher risk of HIV infection. These populations will differ from region to region, according to the epidemiological situation. In most high and middle income countries potential target groups for an initial HIV immunization programme would include intravenous drug users, gay men, commercial sex workers, and high-risk heterosexuals, as well as healthcare workers exposed to blood. In sub-Saharan Africa, future HIV immunization programmes might include larger segments of the population. In order to plan future vaccination programmes it is important to estimate the need (size of target population) and the demand (uptake in target populations) for future HIV vaccines. In addition to the public sector demand for an HIV vaccine (to be used in public health programmes), there will also be a private sector demand driven by the willingness and ability of individuals and employers to pay for the vaccine. HIV vaccines would need to be delivered as part of comprehensive HIV prevention packages, including behavioral and health promotion interventions. This would be especially important with vaccines of moderate efficacy, in order to prevent increased risk behavior among vaccine recipients. To avoid false expectations, the vaccine message would need to be recast as part of the total prevention strategy, rather than the ‘‘magic bullet’’ that people have come to expect. Initial deployment of HIV vaccines could proceed through targeted vaccination campaigns, drawing from experience with other vaccines. These campaigns would be complex and expensive, and would require full participation and collaboration from all levels of the community, as well as considerable strengthening of the infrastructures required for vaccine delivery. Current candidate vaccines in phase III trials may not be appropriate for much of Africa and South Asia, two areas most in need of an HIV vaccine. Credible international efforts (`‘push and pull’’ mechanisms) are needed to create incentives for the industry to develop vaccines for these regions. Feasible financing mechanisms may have to be established to cover the cost of production and delivery of vaccines, in order to ensure equitable access to HIV vaccines around the world. In parallel to the deployment of the initial vaccine, additional bridging studies and effectiveness trials may be needed to expand vaccine use. Research should also continue at an increased pace to develop new generations of more effective vaccines, especially vaccines appropriate to Africa. Achieving these goals will require real political commitment from government and international organizations, to be materialized in specific actions and budget allocations. The daunting challenge of making future effective vaccines accessible to all populations in need will require a sustained collaborative effort on the part of all parties involved.