Response in a bronchodilator must be expressed by the combining the largest amplitude of response with the highest reproducibility. The forced expiratory volume in the first second (FEV1) meets these conditions and thus represents the best test. The bronchodilator response is commonly expressed in percent of the initial FEV1. Through a mathematical bias, this method overestimates the response in the most severely obstructed subjects. Expressing the bronchodilator response in percent of predicted FEV1 eliminates this distortion and must be preferred.