Limited cooperation and low tidal volumes in infants make aerosol therapy difficult. We measured the amount of drug delivered from two baby spacer devices especially developed for use in infants. Designed as a randomized crossover study, aerolized budesonide from a pressurized metered dose inhaler (pMDI) was collected in the inspiratory filter interposed between the face mask and the spacer in 13 infants aged from 2 to 19 months old. The study was performed in connection with pulmonary function testing with a plethysmograph, and the children were sedated with cloral hydrate. Two small-volume baby spacer devices were used: a Babyhaler spacer (GlaxoWellcome, Hertfordshire, UK) made of polycarbonate with a volume of 350 mL and a built-in dead space of 40 mL and a NebuChamber spacer (AstraZeneca, Lund, Sweden) made of stainless steel with a volume of 250 mL and no dead space. Budesonide delivery from the NebuChamber was significantly higher than from the Babyhaler: 38.2% (range, 28.3%-47.5%) of the nominal dose versus 12% (range, 3.3%-21.25%) of the nominal dose of 400 micrograms of budesonide (P = 0.002). The inhaled mass of budesonide from the Babyhaler correlated significantly with skin surface area (r = 0.68, P = 0.018), weight (r = 0.66, P = 0.019), height (r = 0.69; P = 0.017), tidal volume (r = 0.82; P = 0.004), and minute volume (r = 0.67; P = 0.019). No correlations were found between these variables and the inhaled mass of budesonide from the NebuChamber. The results indicate that the design of the NebuChamber spacer affords stable drug delivery in infants and that a large variability in the inhaled mass of drug may be found when infants are inhaling from different baby spacers.