BACKGROUND Currently available nebulisers are inefficient and show variable aerosol deposition in the lung owing to the differences in the particle size generated. The aim of this study was to compare systemic absorption and bronchodilator effects of salbutamol given via a novel ("Ventstream") and a conventional ("Hudson Updraft II") nebuliser system, having initially evaluated the performance of both nebulisers in vitro. The "Ventstream" nebuliser uses a one way valve system to provide an additional inspiratory side flow to match aerosol delivery with tidal volume. METHODS Nebuliser output and particle distribution from 10 Ventstream and 10 Hudson nebulisers were compared in vitro. Eight asthmatic patients, FEV1 55(2)% predicted, were then randomised to receive salbutamol via Ventstream or Hudson nebulisers on separate days. On each day cumulative doses of inhaled salbutamol were given: 1.25 mg, 2.5 mg (1.25 + 1.25 mg), and 5.0 mg (2.5 + 2.5 mg). Airways responses, systemic responses, and plasma salbutamol concentrations were measured at each dose and for up to 240 minutes after the final dose. RESULTS The in vitro comparison showed a greater respirable fraction with a higher percentage volume of particles < 5 microns in diameter using Ventstream than Hudson nebulisers (mean (95% CI) for difference): 25.4% (95% CI 22.4% to 28.3%), and a higher aerosol rate of output: 0.08 g/min (95% CI 0.05 to 0.11 g/min). At the 5.0 mg dose the Ventstream produced a twofold greater concentration of plasma salbutamol than the Hudson nebuliser (AUC0-240): 392.1 ng/ml.min (95% CI 240.6 to 543.6 ng/ml.min). There was a higher AUC0-240 for PEFR with the Ventstream than with the Hudson nebuliser: 74.161 x 10(2) (95% CI 39.50 to 108.821 x 10(2). For FEV1 and FEV25-75 there was a difference in the peak response between the 5.0 mg and 2.5 mg doses with the Ventstream only. Extrapulmonary beta 2 responses were greater with the Ventstream than with the Hudson at 2.5 mg and 5.0 mg doses, although the differences did not appear to be clinically relevant. CONCLUSIONS The Ventstream produced a twofold increase in the delivery of salbutamol to the lung compared with the Hudson nebuliser, and there was an associated increase in systemic beta 2 responses with an improvement in some parameters of bronchodilator efficacy. As a consequence of improved delivery with the Ventstream, it may be possible to halve the drug dose to produce similar bronchodilator efficacy at reduced cost. Further studies are required to evaluate the value of the Ventstream for delivery of nebulised antibiotics and corticosteroids.