PURPOSE:To determine whether people with chronic obstructive pulmonary disease (COPD) have characteristics that predict a clinically meaningful response to pulmonary rehabilitation (PR) that includes an education component compared with exercise training alone. METHODS:Participants were classified as responders or nonresponders to 2 models of PR; exercise training and education (ET + ED, n = 113) or exercise training alone (ET, n = 85). Responders were defined as those who achieved a clinically meaningful change in 6-min walk distance (6MWD) or any of the 4 domains of the Chronic Respiratory Questionnaire (CRQ). Baseline characteristics were compared between responders and nonresponders. The associations between baseline data and change in 6MWD and CRQ following PR were examined, and a binary logistic regression analysis was conducted for each model and primary outcome. RESULTS:There were no significant differences between the PR models in proportion of responders (ET + ED 92% vs ET 93%). Lower baseline CRQ scores predicted response in respective CRQ domains for fatigue, emotion, and mastery in the ET + ED group, and for dyspnea, fatigue, and mastery in the ET group. Higher baseline self-efficacy predicted 6MWD response and higher socioeconomic status predicted response in CRQ fatigue in the ET model only. There was no predictor of 6MWD response in the ET + ED group. CONCLUSIONS:Baseline characteristics did not reliably predict a clinically meaningful response to PR that included education. For exercise training alone, higher self-efficacy was a significant predictor for greater improvements in 6MWD, suggesting that those with higher confidence should be considered for this model.