We tested the ability of chest physicians to detect increased airways responsiveness (IAR) in patients presenting with symptoms suggestive of asthma. Physicians ordering bronchial provocation tests for diagnostic purposes were asked to predict the clinical probability (based on the patient's history, examination, and spirometry) of IAR (defined by a PC20 for FEV1 less than or equal to 8 mg/ml of histamine). Presenting symptoms in the 38 subjects aged 16 to 78 years included cough, wheeze, dyspnea, chest tightness, and recurrent colds. There was no association between IAR and clinical probability dichotomized as less than or equal to 50 percent or greater than 50 percent (p = .16, Fisher test). The association was no stronger using only subjects in whom the clinical prediction was greater than 67 percent or less than 33 percent. If knowledge of the level of airway reactivity would influence subsequent patient management (eg, prescription of antiasthmatic therapy), then its clinical prediction should not be relied upon.