BACKGROUND AND OBJECTIVE: Pulmonary rehabilitation (PR) is a cornerstone of care in chronic respiratory diseases; yet its benefits diminish over time. Repeating PR may be beneficial; however, little is known about the characteristics and outcomes of repeaters. This study aimed to establish the proportion of repeaters, identify characteristics that predict repetition and compare the magnitude of benefits achieved between initial and subsequent programmes. METHODS: Patients with stable chronic respiratory diseases who attended PR over a 9-year period were included. Outcome measures included the 6-min walk distance (6MWD) and the Chronic Respiratory Disease Questionnaire-Self-Reported (CRDQ-SR). Independent predictors of repeating were identified. RESULTS: Of 296 patients, 59 (20%) repeated PR, most within 1-3 years. Following the initial programme, repeaters had significant decline in 6MWD (-96.1 ± 84.6 m; P < 0.001) and CRDQ-SR scores (mean decline -3.6 points, range -0.1 to -7.9 points; P < 0.005). The improvement in 6MWD was less in the repeat programme compared with the first (38.4 ± 50.7 m vs 67 ± 40.4 m; P = 0.005), while the change in CRDQ-SR was similar in all domains. A chronic obstructive pulmonary disease diagnosis increased the odds of repeating PR (odds ratio (OR) 4.8; P = 0.005) while improved mastery in the initial programme reduced the odds (OR 0.9; P = 0.033). CONCLUSIONS: One in five patients repeated PR, achieving clinically significant improvements in exercise tolerance and quality of life. Patients with small improvements in disease mastery after initial PR were more likely to repeat the programme and may benefit from earlier intervention or longer duration PR.