Background: Asthma is generally accepted as an inflammatory disease that needs steroid treatment. However, when to start with inhaled steroids remains unclear. A study was undertaken to determine when inhaled corticosteroids should be introduced as the first treatment step. Objective: To investigate the effectiveness of early introduction of inhaled steroids on decline in lung function in steroid-naïve subjects with a rapid decline in lung function in general practice. Subjects: Patients with signs/symptoms suspect of asthma (i.e., persistent and/or recurrent respiratory symptoms) and a decline in forced expiratory volume in 1 s (FEV1) during 1-year monitoring of 0.080 l or more and reversible obstruction (≥10% predicted) or bronchial hyperresponsiveness (PC20≤8 mg/ml) were studied. They had been identified in a population screening aiming to detect subjects at risk for chronic obstructive pulmonary disease (COPD) or asthma. Design: A placebo-controlled, randomized, double-blind study. Methods: 75 subjects out of a random population of 1155 were found eligible, and 45 were willingly to participate. Subjects were randomly treated with placebo or fluticasone propionate 250 µg b.i.d., and FEV1 and PC20 were monitored over a 2-year period. Outcome variables: The primary outcome measure was decline in FEV1; the secondary outcome measure was bronchial hyperresponsiveness (PC20). Results: 22 subjects were randomly allocated to the active group with inhaled corticosteroids and 23 to placebo. Change of FEV1 in the active treated group was +43 ml in post-bronchodilator FEV1 (p =0.341) and +62 ml/year (p =0.237) in pre-bronchodilator FEV1 after 1 year, and −22 ml (p =0.304) for post-bronchodilator FEV1 and −9.4 ml (p =0.691) for pre-bronchodilator FEV1 after 2 years, compared to placebo. The effect on PC20 was almost one dose-step (p =0.627) after 1 year and one dose-step (p =0.989) after 2 years.Conclusion: In this study, the early introduction of inhaled corticosteroids in newly diagnosed asthmatic subjects with rapid decline in lung function did not prove to be either clinically relevant or statistically significant in reversing the decline in FEV1. For PC20, no significant changes were detected.