The hypothesis that exercise limitation of respiratory origin can be predicted accurately from the lung function impairment has been tested using maximal oxygen uptake (_O2max) as the dependent variable in a multiple regression analysis. The subjects were 157 men who met objective criteria for exercise being limited by respiratory impairment. _O2max (mean value, 1.38 L min⁻¹) was described by FEV₁ and single-breath lung transfer factor (diffusing capacity) for carbon monoxide (Tl′) singly or in combination, but the accuracy was poor (at best, standard error of the estimate, 0.36 L min⁻¹; r², 29.1%). FEV₁ could be replaced by FVC and FEV₁/FVC. Description of _O2max was improved by also including in the equation the variables age, fat-free mass, and submaximal exercise ventilation (e). Transfer factor did not then contribute significantly. _O2max as percent of predicted (mean value of 60%) was described by %FVC or %FEV₁, but the accuracy was poor (SEE, 16.0%; r², 14%). Prediction was improved somewhat by the alternative use of inspiratory vital capacity and FEV₁/FVC. Transfer factor did not contribute additional information; however, inclusion of e materially improved the accuracy (SEE, 12.9%; r², 44%). Among a subgroup of 35 men whose lung disease was due to asbestos, %Tl′ or transfer factor measured using a multibreath estimate of residual volume (%Tl_CO) made a small contribution to the explained variance, e.g.: %_O2max = 0.440/0 FEV₁ −0.78 e + 0.16% Tl_CO + 52.3 SEE 7.27%. This equation also described the %_O2max of all subjects (SEE, 13%). It was concluded that in patients whose exercise capacity was limited by respiratory factors, the limitation could not be assessed accurately in terms of FEV₁, FVC, and transfer factor. These indices were not independent, and their predictive power was poor. for the accuracy to be acceptable, additional indices were needed; for example, ventilation at a standard oxygen uptake during submaximal exercise.