Introduction: Fatigue is a prominent symptom in chronic obstructive pulmonary disease (COPD) and it has distinctive features; however, there is a need for a robust scale to measure fatigue in COPD. Methods: At baseline, 122 patients with COPD (forced expiratory volume in 1 s (FEV1) 52%, women 38%, mean age 66 years) completed a pilot fatigue scale covering a pool of 57 items and underwent a range of tests, including indicators of mood and a short general fatigue questionnaire. All patients responded to the 57-item scale and it was readministered to a subset of 30 patients. The pilot scale was first subjected to constructive validated shortening steps and then to a principal components analysis. Results: The Manchester COPD fatigue scale (MCFS) consists of 27 items, loading into three dimensions: physical, cognitive and psychosocial fatigue. Internal consistency (Cronbach’s α = 0.97) and test–retest repeatability (r = 0.97, p<0.001) were tested. It had significant convergent validity, correlating with the FACIT (Functional Assessment of Chronic Illness Therapy) fatigue scale and the fatigue in Borg scale at baseline and after a 6 minute walk distance (6MWD) test (r = −0.81, 0.53 and 0.63, respectively, p<0.001). Its scores were associated with BODE, SGRQ (St George’s Respiratory Questionnaire) and MRC (Medical Research Council) dyspnoea scores (r = 0.46, 0.8 and 0.51, respectively, p<0.001). The scale demonstrated meaningful discriminating ability; patients who walked <350 m in a 6MWD test as well as depressed patients (⩾16 scores in the Center for Epidemiologic Study on Depression (CES-D) scale) had nearly twice as high fatigue scores as those who walked ⩾350 m or were not depressed (p<0.001). Conclusion: The MCFS provides a simple, reliable and valid measurement of total and dimensional fatigue in moderate stable COPD.