While a number of studies remain to be completed, evidence is mounting that there is no demonstrable risk for infants immunized with vaccines containing thiomersal. Epidemiological studies in the US have shown no developmental or other central nervous system abnormalities resulting from exposure to vaccines containing thiomersal. During the initial evaluation of thiomersal in vaccines during 1999, the toxicological profile of ethyl mercury was unknown and presumed to be the same as that of methyl mercury. Enough evidence has accumulated since then to indicate the profiles of the two compounds are different in crucial aspects. To date, one study has measured blood levels of total mercury in vaccinated infants and reports only a brief low-level exposure with rapid excretion of mercury. It is not yet known for sure how much (if any) vaccine-derived ethyl mercury in the blood crosses the blood-brain barrier. For the most part, the use of thiomersal as a vaccine preservative has been convincingly shown to be safe. The scientific evidence is not yet sufficiently strong to provide the same level of assurance for thiomersal-containing vaccines for use in pregnant women or the premature or low birth weight infant. There is an increased sensitivity of the fetal brain to mercury whether it is ethyl or methyl mercury. While there is no evidence to support the contention, it is at least theoretically possible that very low birth weight premature infants may be at increased risk from thiomersal-containing vaccines. Until such time as the scientific evidence is to hand, thiomersal-free presentations of hepatitis B are to be preferred for the birth dose. Given the same levels of exposure, adults are at much lower levels of risk because of increased body mass. It is not possible to prove that thiomersal is completely safe-epidemiology can only quantify a risk, not prove its absence.