Pressurized metered dose inhalers (pMDI) are widely used together with spacers for the treatment of asthma in children. However, the variability of daily medication dose for pMDI/spacer combinations is not known. Electrostatic charge is a potential source of dose variability. Metal spacers have no static charge. This study assessed and compared within-subject variability of aerosol delivery of metal and plastic spacers. This was a randomized, crossover study in children with stable asthma aged 1-4 (group I, n=17) and 5-8 (group II, n=16) yrs. In both groups the amount of drug delivered to the mouth by a metal spacer (Nebuchamber) and one of two plastic (polycarbonate) spacers, i.e. Babyhaler in group I and Volumatic in group II was measured. The metal and plastic spacers were tested at home in a randomized order for 7 days each, using budesonide (200 microg b.i.d.). Aerosol was collected on a filter positioned between spacer and facemask or mouth. Budesonide on the filter was assessed by high performance liquid chromatography. The mean filter dose for each child (mean+/-SD) during the 7 days was expressed as a percentage of the nominal dose. Within-subject variability was expressed as coefficient of variation (CV). Mean filter dose in group I was 41.7+/-10.1% for Nebuchamber and 26.0+/-4.0% for Babyhaler (p<0.001). Mean filter dose in group II was 50.2+/-9.2% for Nebuchamber and 19.4+/-7.2% for Volumatic (p<0.001). Mean CV in group I was 34% for Nebuchamber and 37% for Babyhaler (p=0.44). Mean CV in group II was 23% for Nebuchamber and 34% for Volumatic (p=0.003). There was substantial within-subject dose variability in aerosol delivery in children using a pMDI/spacer at home. This variability was lower for the metal than for the plastic spacer in children 5-8 yrs of age. The dose delivered to the mouth was about two-fold higher for the metal than the plastic spacer independent of age.