Avoidable deaths from asthma continue. Some of these result from the difficulty in determining the severity of an acute asthma attack at the initial assessment. This study evaluated the relation of pulse oximetry with other markers of severity in 46 patients attending an accident and emergency (A&E) department with acute asthma. Neither oxygen saturation nor peak flow correlated with length of admission or with other "retrospective" markers of severity. Attempts to collect follow up data (for example, peak flow charts) from patients discharged from the A&E department failed. It proved impossible to determine whether pulse oximetry predicts which adults can be discharged.