The best means for optimal delivery of drugs into lungs of infants with bronchopulmonary dysplasia (BPD) is uncertain. We aimed to measure radio-aerosol deposition of salbutamol by jet nebulizer and metered dose inhalers (MDI) in ventilated and non-ventilated BPD infants. In a randomized, crossover sequence, salbutamol lung deposition was measured using an MDI (2 puffs or 200 micrograms) or sidestream jet nebulizer (5 minutes of nebulization with 100 micrograms/kg) in 10 ventilated (mean birthweight, 1,101 g) and 13 non-ventilated (mean birthweight, 1,093 g) prematurely born infants. Non-ventilated infants inhaled aerosol through a face mask, connected to a nebulizer or an MDI and spacer (Aerochamber). Ventilated infants received aerosol from an MDI + MV15 Aerochamber or a nebulizer inserted in the ventilator circuit. Lung deposition by both methods was low: mean (SEM) from the MDI was 0.67 (0.17)% of the actuated dose, and from the nebulizer it was 1.74 (0.21)% and 0.28 (0.04)% of the nebulized and initial reservoir doses, respectively. Corresponding figures for the ventilated infants were 0.98 (0.19)% from the MDI and 0.95 (0.23)% and 0.22 (0.08)% from the nebulizer. In both groups, and for both methods of delivery, there was marked inter-subject variability in lung deposition and a tendency for the aerosol to be distributed to the central lung regions.