A number of affluent countries are moving to eliminate thiomersal (thimerosal), an ethylmercury preservative, from vaccines as a precautionary measure because of concerns about the potential adverse effects of mercury in infants. The WHO advocates continued use of thiomersal-containing vaccines in developing countries because of their effectiveness, safety, low cost, wide availability and logistical suitability in this setting. The guidelines for long-term mercury exposure should not be used for evaluating risk from intermittent single day exposures, such as immunisation using thiomersal-containing vaccines. Similar or higher mercury exposures likely occur from breast feeding and the health benefit of eliminating thiomersal from a vaccine, if any, is likely to be very small. On the other hand, the benefits accrued from the use of thiomersal-containing vaccines are considerably greater but vary substantially between affluent and developing regions of the world. Because of the contribution to overall mercury exposure from breast milk and diet in later life, the removal of thiomersal from vaccines would produce no more than a 50% reduction of mercury exposure in infancy and <1% reduction over a lifetime. Different public policy decisions are appropriate in different settings to achieve the lowest net risk, viewed from the perspectives of the individual vaccinee or on a population basis. In developing regions of the world, at least over the next decade, far more benefit will accrue from protecting children against widely prevalent vaccine-preventable diseases by focusing efforts aimed at improving infant immunisation uptake by using current, inexpensive, domestically-manufactured, thiomersal-containing vaccines, than by investing in thiomersal-free alternatives.