H1-receptor antagonists are considered central to the treatment of atopic dermatitis (AD)-associated pruritus and are widely used in the treatment of AD despite a lack of double-blind, randomized clinical trials. In this study we analyzed the effects of the long-term use of cetirizine on the severity, natural history, and treatment of AD. In the prospective, multi-country, double-blind, randomized, placebo-controlled Early Treatment of the Atopic Child (ETAC) study, 817 infants (12-24 months of age) who suffered from AD at study entry and with a history of atopic disease in a parent or sibling were treated for 18 months with either cetirizine (0.25 mg/kg) or placebo twice daily. All concomitant medications for the treatment of AD were allowed but had to be recorded by the investigator in the case report form; the concomitant use of H1-antihistamines was discouraged. The primary end-point for efficacy was the onset of asthma. Secondary parameters of efficacy, however, were the consumption of concomitant medications for AD (topical and systemic treatment) and the severity of symptoms related to AD, which was rated with the AD scale, SCORAD. The severity of AD, as measured by SCORAD, decreased significantly (p < 0.001) over the study period (18 months) in both groups. Other oral H1-antihistamines were significantly more often used in the placebo group than in the cetirizine group (24.9% vs. 18.6%, p = 0.03). The number of infants who developed urticaria during the study period was significantly lower with cetirizine treatment (placebo group: 16.2%; cetirizine group: 5.8%; p < 0.001). For the treatment of AD, mild topical corticosteroids (class I, e.g. hydrocortisone) were used in 41.6% of the patients (placebo group 41.6%, cetirizine group 41.7%) and moderate-to-potent topical corticosteroids (class II, III, IV) in 55.0% (respectively 56.4% and 53.5%). The duration of the use of topical moderate-to-potent corticosteroids differed between the cetirizine group and the placebo group (mean percentage of days: placebo 25.2, median 2.4; cetirizine mean 18.8, median 0.95 p = 0.067, Mann-Whitney test, not statistically significant). In sub-groups of infants with a SCORAD of >or= 25, this cortico-sparing effect was statistically significant (placebo 35.1 vs. 25.8 in the cetirizine group; p = 0.014, Mann-Whitney test). In conclusion, in view of the proven safety of cetirizine, the use of this drug might help to reduce the duration and the amount of moderate-to-strong topical corticosteroids used in the treatment of infants and children with AD. However, further studies designed with the primary end-point of AD are clearly indicated to confirm the benefits of the use of H1-antihistamines in the management of AD.