This study aimed at identifying in a daily-life setting the influence of facemask design on drug delivery via a spacer to young children. In a 4-week randomized crossover study, 24 children (7-23-months old) with recurrent wheeze tested the AstraZeneca, Galemed, and Hans Rudolph facemask combined with the NebuChamber at home. Each mask was tested twice daily for seven consecutive days. Filters positioned between the NebuChamber and facemask trapped the budesonide aerosol (200 microg, Pulmicort). Parents were asked to score the child's degree of cooperation during administration on diary cards. The administration procedure was evaluated through video recordings. Mean filter dose (standard deviation (s.d.)), expressed as % of nominal dose, was 39% (14), 47% (12), and 42% (11) for the AstraZeneca, the Galemed and the Hans Rudolph mask, respectively. Irrespective of the degree of cooperation, the Galemed mask gave significantly higher mean filter doses than the other masks (level of significance) (p < 0.045). Median (range) within-subject dose variability, expressed, as coefficient of variation (CV), was 37% (19-255), 32% (9-114), and 30% (9-115) for the AstraZeneca mask, the Galemed mask and the Hans Rudolph mask, respectively, not significant. Dose variability increased with decreasing cooperation for all three masks (p = 0.007). Drug delivery to young children with recurrent wheeze by means of the NebuChamber can be enhanced using the Galemed facemask. Dose variability seems to be independent of facemask design but mainly depends on cooperation.