To investigate whether the obesity survival paradox, where obesity is associated with improved survival, exists for pneumonia supported with extracorporeal membrane oxygenation (ECMO). Between January 2014 and December 2015, 223 patients with acute respiratory failure who underwent ECMO in 11 hospitals in South Korea were enrolled retrospectively, and data relating to pneumonia cases were analyzed. Patients were divided into groups according to their pre-treatment body mass index (BMI): obese (BMI ≥ 25 kg/m2) and non-obese (BMI < 25 kg/m2). The BMI cut-off was adopted from the World Health Organization for Asian populations. In total, 84 patients had pneumonia: obese group, 26; non-obese group, 58 (mean BMI, 27.8 vs. 21.4, p < 0.001). Pre-ECMO parameters and rescue therapy did not differ between the groups. The rate of successful weaning from ECMO was not different between the groups (76.9% vs. 72.4%, p = 0.753), but survival-to-discharge (73.1% vs. 36.2%, p = 0.002) and 6-month survival (69.2% vs. 36.2%, p = 0.005) rates were higher in the obese group. After adjusting for age and the pre-treatment PaO2/FiO2 ratio, low BMI was significantly associated with 6-month mortality (odds ratio 3.28, 95% confidence interval: 1.06–11.03, p = 0.044). An obesity survival paradox exists in pneumonia supported with ECMO. •Successful weaning of ECMO in obese patients is associated with improved outcome when compared to non-obese patients.•Obesity should not limit the use of ECMO.•Pneumonia patients requiring ECMO experience significant weight loss.