A pear shaped non-electrostatic spacer, composed of steel with a volume of 250 ml and equipped with a facemask containing integrated inlet and outlet valves for inspiration and expiration, was compared with three plastic spacers. The plastic spacers were primed with repeated puffs from a budesonide pressurised metered dose inhaler (p-MDI) to minimise the electrostatic charge on the plastic. The procedure prolonged the half life (t1/2) of the aerosol in the Nebuhaler from nine to 32 seconds. A normal cleaning procedure reduced the aerosol t1/2 back to baseline. The t1/2 of the aerosol in the metal spacer was 27 seconds and independent of the use of p-MDI. In vitro the maximum dose of budesonide from a p-MDI, expressed as a percentage of the nominal dose, was 56% from the non-electrostatic spacer, 61% from the Nebuhaler, 45% from the Babyhaler, and 30% from the AeroChamber. In 124 children, age 6 months to 6 years, suspected to have asthma the non-electrostatic spacer delivered a mean total dose of budesonide aerosol of 39% of the nominal dose, which was significantly higher than the Babyhaler (28%), the Nebuhaler (21%), and the AeroChamber (19%). These differences were most pronounced in children younger than 4 years. The improved dose delivery from the small volume non-electrostatic spacer is probably related to the non-electrostatic spacer material and the valves which assured unidirectional airflow from the spacer without adding any dead space in the inspiratory channel. The non-electro-static spacer should improve the cost effectiveness of aerosol treatment and, as the counteracting effects of proming and recharging of the plastic from cleaning are avoided, should deliver a more reliable dose.