The importance of reversible airflow obstruction to the prognosis of asthma and chronic obstructive pulmonary disease (COPD) is not clear. We tested the hypothesis that reversibility to corticosteroid and bronchodilator is not an independent predictor of prognosis, but merely reflects a component of the maximal attainable lung function, which is the best spirometric predictor of survival. During a 6-yr period (1983-1988), 1,586 subjects with asthma or COPD underwent standardized bronchodilator and corticosteroid reversibility tests at a chest clinic in Copenhagen. The vital status was obtained by September 1997. The relationship between mortality and age, gender, smoking, FEV1, and reversibilities was examined by Cox proportional hazards analyses. Of 1,586 subjects, 850 had died before September 1997. Age, smoking, and FEV1 were significant predictors of mortality. After controlling for baseline FEV1, bronchodilator and corticosteroid reversibility were significantly associated with better survival. However, after controlling for best FEV1 all reversibilities became nonsignificant and nonpredictive. The combined use of corticosteroid and bronchodilator reversibility in survival analyses is a novel approach, and we have shown that both contribute to survival prediction to the extent that they modify FEV1. However, reversibility per se does not influence survival in subjects with moderate to severe asthma or COPD.