Abstract Introduction. The cardiopulmonary exercise test (CPET) and the 6-minute walk test (6MWT) are used to prescribe the appropriate training load for cycling and walking exercise in patients with chronic obstructive pulmonary disease (COPD). The primary aims were: (i) to compare estimated peak work rate (Wpeakestimated) derived from six existing Wpeak regression equations with actual peak work rate (Wpeakactual); and (ii) to derive a new Wpeak regression equation using six-minute walk distance (6MWD) and conventional outcome measures in COPD patients. Methods. In 2906 patients with COPD, existing Wpeak regression equations were used to estimate Wpeak using 6MWD and a new equation was derived after a stepwise multiple regression analysis. Results. The 6 existing Wpeak regression equations were inaccurate to predict Wpeakactual in 82% of the COPD patients. The new Wpeak regression equation differed less between Wpeakestimated and Wpeakactual compared to existing models. Still, in 74% of COPD patients Wpeakestimated and Wpeakactual differed more than (+/-) 5 watts. Conclusion. In conclusion, estimating peak work load from 6MWD in COPD is inaccurate. We recommend assessment of Wpeak using CPET during pre-rehabilitation assessment in addition to 6MWT.