Sixteen Chinese patients with chronic fixed airflow obstruction (mean age 62.5 years; mean forced expiratory volume in 1 second (FEV1) of 0.85 litres, and mean ratio FEV1/forced vital capacity (FVC) of 0.45) entered a randomized double-blind crossover trial comparing prednisolone 40 mg orally daily for 2 weeks with placebo. Nine of the 16 patients (56%) had an increase in FEV1 of 15% or more after prednisolone. Compared with placebo, prednisolone significantly improved objective measurements (mean FEV1 by 21.4%, mean FVC by 11.9%, mean daily peak expiratory flow rate by 22.6%), subjective measurement (mean dyspnoea score by 16%) and exercise performance assessed by the distance walked in 12 minutes (12MD, by 5.5%). Objective improvements in FEV1 and PEFR were correlated with subjective upgrading of the dyspnoea score, but were not paralleled by an improvement in 12MD. Changes in FVC correlated with neither. FEV1 was thus the best index of objective measurement. The following characteristics were examined for reliability in predicting steroid responsiveness: variability of symptoms, wheezing score, prick skin test positivity, initial bronchodilator response, peripheral blood and sputum eosinophilia, and serum and sputum immunoglobulin E (IgE) level. None was found to be useful. A therapeutic trial of oral corticosteroid should be considered in patients with chronic fixed airflow obstruction on an individual basis.