OBJECTIVE To assess the relationship of distance ambulated during the 6-min walk test (6'WT) to maximal oxygen consumption (VO2 max). METHODS Multivariate analysis of patient characteristics to VO2 max. METHODS Pre-lung transplant evaluation. METHODS 60 patients (22 men, 38 women; mean age, 44 years) with end-stage lung disease (mean FEV1 and forced vital capacity of 0.97 and 1.93, respectively). RESULTS The 6'WT was performed on a level hallway surface, and VO2 max was obtained during maximal cycle ergometry exercise testing with respiratory gas analysis. Multivariate analysis of patient characteristics (age, sex, weight, FEV1, FVC, diffusing capacity for carbon monoxide (DCO), 6'WT distance ambulated, number of rests per 6'WT, and the maximal heart rate, blood pressure, rate-pressure product, respiratory rate, oxygen saturation, rating of perceived exertion, and amount of supplemental oxygen used during the 6'WT) was performed on two groups of 30 patients each (group A or B) who were randomly assigned to either group by a process of random selection using a computer-generated random numbers program. Distance ambulated was the strongest independent predictor of VO2 max (r = 0.73; p < 0.0001) in both groups, and adding age, weight, and pulmonary function test results (FVC, FEV1, and DCO) to the regression equation increased the correlation coefficient to 0.83. Because of the significant correlation of distance ambulated during the 6'WT to VO2 max, the prediction equation obtained from the multivariate analysis of group A, VO2 max = 0.006 x distance (feet) +3.38, was used to estimate the VO2 max of the group B patients. No significant difference was observed between the estimated (x +/- SD = 8.9 +/- 2.4 mL/kg/min) and observed (x +/- SD = 9.4 +/- 3.8 mL/kg/min) VO2 max (mean difference, 0.5 mL/kg/min; SD of the difference = 2.88). CONCLUSIONS The distance ambulated during a 6'WT can predict VO2 max in patients with end-stage lung disease. The addition of several patient characteristics can increase the ability to predict VO2 max and account for more of the variability. Such information is valuable when assessing patient response to therapeutic intervention if respiratory gas analysis is unavailable or impractical.