The primary purpose of this study was to determine if there are three distinct factors representing ratings of dyspnea during laboratory exercise, clinical ratings of dyspnea, and pulmonary function in patients with moderate to severe chronic obstructive pulmonary disease (COPD) (n = 92) using factor analysis. Subjects (mean age 66 +/- 7 yrs; FEV1% predicted 44.7 +/- 14.0) were randomly assigned to one of three treatment groups to test the effects of three education and exercise training programs. Outcomes were evaluated at baseline and at 2 months after the intervention. Dyspnea ratings with laboratory exercise (SOB) were measured during incremental (ITT) and endurance (ETT) treadmill tests, and a six-minute walk (6MW) using the modified Borg scale. Clinical measures of dyspnea were measured with the Baseline and Transitional Dyspnea Index (BDI/TDI), UCSD Shortness of Breath Questionnaire (SOBQ), Medical Research Council Dyspnea Scale (MRC), Dyspnea subscale of the Chronic Respiratory Questionnaire (CRQ-D), and a global dyspnea question. Pulmonary function parameters included FEV1% predicted, FEV1/FVC, and RV/TLC. The factor analysis yielded three factors that accounted for 58.7% of the total variance in the data: Factor 1, "Dyspnea with Laboratory Exercise" comprised ETT SOB end, ETT SOB isotime, ITTSOB end, ITT SOB isotime, ITT SOB/Time and ETT SOB/Time. Factor 2, "Clinical Dyspnea," comprised 6MW SOB, 6MW SOB/Feet, BDI, SOBQ, MRC, Global SOB, CRQ-D. Measures of airway resistance (FEV1% predicted, FEV1/FVC) and hyperinflation (RV/TLC) loaded on a third factor, "Pulmonary Function." An additional post hoc factor analysis with post-intervention data provided similar results. The Global SOB question and ITT SOB isostage variables were relatively more sensitive to change compared to the other outcome variables. We conclude that pulmonary function, clinical ratings of dyspnea, and laboratory ratings of dyspnea are three separate and independent factors and should be included in the routine clinical evaluation of patients with COPD.