A characteristic relationship between infection and host age, with lev
els of infection reaching a peak in particular age classes, has been r
eported for many parasites. However, several field studies have shown
that this relationship is not invariant: if age-infection data are com
pared across host populations, the peak level of infection is higher a
nd occurs at a younger age when the transmission rate is high, and is
lower and occurs at an older age when if is low. This pattern is calle
d the 'peak shift'. Here, Mark Woolhouse reviews the evidence or and t
he implications of the peak shift. The peak shift is consistent with t
he predictions of mathematical models that assume gradually acquired p
rotective immunity, and this interpretation is supported by experiment
al studies using animals. This agreement between theory, experimental
evidence and field studies strongly suggests that acquired immunity ha
s a major impact on epidemiological patterns Mot only for parasites su
ch as malaria, where the importance of acquired immunity is not in dou
bt, but also for many parasitic helminths, where the role of acquired
immunity is less widely accepted.