The accidental release of chlorine gas is a constant threat in urban areas. The purpose of this randomized, blinded, controlled experiment was to examine the effects of post-injury administration of inhaled or intravenous corticosteroid in chlorine gas-injured pigs followed for 23 h. Anaesthetized, ventilated pigs (n = 24) in the prone position were exposed to chlorine gas (400 parts per million in air) (1160 mg/m3) for 15 min, then randomly allocated to receive inhaled budesonide (BUD) and intravenous placebo, intravenous betamethasone (BETA) and inhaled placebo or inhaled and intravenous placebo. Haemodynamics, gas exchange and lung mechanics were evaluated for 23 h after exposure to chlorine gas. Airway and pulmonary artery pressures increased and arterial oxygenation fell sharply (from 13.5 +/- 0.8 to 6.7 +/- 0.9 kPa, P < 0.001) after chlorine gas exposure. These immediate changes were followed by a gradual improvement over 5-7 h to a stable level of dysfunction for the rest of the experiment in placebo animals. Arterial oxygen tension, pulmonary vascular resistance and airway pressure recovered faster and more completely in the budesonide and betamethasone groups than in the placebo group (P < 0.01). Lung wet weight to dry weight ratios were greater in the placebo group than in the budesonide and betamethasone groups (6.34 +/- 0.59 vs. 5.56 +/- 0.38 and 5.53 +/- 0.54, respectively, P < 0.05). There was a trend towards lower histological injury scores compared with placebo in animals that received budesonide (P = 0.05) or betamethasone (P = 0.07). Treatment of chlorine gas lung injury with nebulized budesonide or intravenous betamethasone had similar positive effects on recovery of lung function.