The IgE and IgG responses to pertussis toxin were measured in blood samples from 70 children (age 1.5-2.9 years) after primary immunisation with either a non-aluminium adsorbed, whole cell vaccine (n = 34) or an aluminium adsorbed whole cell vaccine (n = 36). Two years later, they received a booster immunisation with either the non-adsorbed (n = 24) or the aluminium adsorbed vaccine (n = 14). Neutralising antibodies to pertussis toxin were higher (P < 0.05) after the three priming doses of the adsorbed vaccine than of the non-adsorbed vaccine, although both groups showed > 90% seropositives after the third dose. IgE antibodies to PT (PT-IgE) were detected in samples from 11/52 children after completed primary immunisation and the levels were low (median < or = 0.1 PRU ml-1) in both groups. No significant differences between the groups were found. PT-IgE levels did not increase after the booster injection. Thus, the aluminium content of the whole cell vaccines influenced the IgG response but not the IgE responses to pertussis toxin. The high rates of PT-IgE responses noted after a booster dose of acellular or whole cell pertussis vaccine to children primed with acellular vaccine in previous studies can therefore be mainly ascribed to the nature of the priming vaccine rather than the aluminium adjuvant.