Hospitalized, malnourished older adults with chronic obstructive pulmonary disease (COPD) have an elevated risk of readmission and mortality. Post-hoc, sub-group analysis from the NOURISH study cohort examined the effect of a high-protein oral nutritional supplement (ONS) containing HMB (HP-HMB) in malnourished, hospitalized older adults with COPD and to identify predictors of outcomes. The NOURISH study (n = 652) was a multicenter, randomized, placebo-controlled, double-blind trial. The COPD subgroup (n = 214) included hospitalized, malnourished (based on Subjective Global Assessment), older adults (≥65 y), with admission diagnosis of COPD who received either standard-of-care plus HP-HMB (n = 109) or standard-of-care and a placebo supplement (n = 105) prescribed 2 servings/day from within 3 days of hospital admission (baseline) and up to 90 days after discharge. The primary study outcome was a composite endpoint of incidence of death or non-elective readmission up to 90-day post-discharge, while secondary endpoints included changes in hand-grip strength, body weight, and nutritional biomarkers over time. Categorical outcomes were analyzed using Cochran-Mantel-Haenszel tests, longitudinal data by repeated measures analysis of covariance; and changes from baseline by analysis of covariance. p-values ≤ 0.05 were considered statistically significant. Multivariate logistic regression was used to model predictors of the primary outcome and components. In patients with COPD, 30, 60, and 90-day hospital readmission rate did not differ, but in contrast, 30, 60, and 90-day mortality risk was approximately 71% lower with HP-HMB supplementation relative to placebo (1.83%, 2.75%, 2.75% vs. 6.67%, 9.52% and 10.48%, p = 0.0395, 0.0193, 0.0113, resp.). In patients with COPD, compared to placebo, intake of HP-HMB resulted in a significant increase in handgrip strength (+1.56 kg vs. -0.34 kg, p = 0.0413) from discharge to day 30; increased body weight from baseline to hospital discharge (0.66 kg vs. -0.01 kg, p < 0.05) and, improvements in blood nutritional biomarker concentrations. The multivariate logistic regression predictors of the death, readmission or composite endpoints in these COPD patients showed that participants who were severely malnourished (p = 0.0191) and had a Glasgow prognostic score (GPS) Score of 1 or 2 had statistically significant odds of readmission or death (p = 0.0227). Among malnourished, hospitalized patients with COPD, supplementation with HP-HMB was associated with a markedly decreased mortality risk, and improved handgrip strength, body weight, and nutritional biomarkers within a 90-day period after hospital discharge. This post-hoc, subgroup analysis highlights the importance of early identification of nutritional risk and administration of high-protein ONS in older, malnourished patients with COPD after hospital admission and continuing after hospital discharge. Copyright © 2020 Elsevier Ltd and European Society for Clinical Nutrition and Metabolism. All rights reserved.