Objectives Our aim was to study the associations between resting lung hyperinflation, weight-bearing exercise-induced dyspnea and adipose distribution in obese and normal-weight COPD patients. Methods We performed a comparison between 80 obese COPD patients (COPDOB) with 80 age- and FEV1 matched normal-weight COPD patients (COPDNW). Dyspnea was assessed by the mMRC scale and the Borg dyspnea score before and after a 6 min walk test. Further characterization included spirometry, body plethysmography and metronome paced tachypnea (MPT) to estimate dynamic hyperinflation. Body composition was assessed with bioelectrical impedance analysis. Associations between dyspnea scores and BMI and body composition groups were studied using logistic regression models. Results COPDOB patients had attenuated increases in TLC, FRC and RV compared to COPDNW patients ( p < 0.01). The groups had comparable 6 min walking distance and ΔFRC upon MPT ( p > 0.05). Compared to COPDNW, COPDOB patients reported more often a mMRC ≥ 2 (65 vs 46%; p = 0.02; OR 3.0, 95% CI 1.4–6.2, p < 0.01) and had higher ΔBorg upon 6MWT: 2.0 (SEM 0.20) vs. 1.4 (SEM 0.16), p = 0.01; OR for ΔBorg ≥ 2: 2.4, 95% CI 1.1–5.2, p = 0.03. Additional logistic regression analyses on the associations between body composition and dyspnea indicated that increased body fat percentage, fat mass index and waist-to-hip ratio were associated with higher ORs for mMRC ≥ 2 and ΔBorg upon 6MWT ≥ 2. Conclusion Despite its beneficial effect on resting lung hyperinflation, adiposity is associated with increased weight-bearing exercise-induced dyspnea in COPD.