Rationale Findings from individual trials of physical rehabilitation interventions in critically ill adults have limited potential for meta-analysis and informing clinical decision-making due to heterogeneity in selection and reporting of outcomes used for evaluation. Objective The objective of this study was to determine a core outcome set (COS) for use in all future trials evaluating physical rehabilitation interventions delivered across the critical illness continuum of recovery. Methods An international, two-round, online, modified Delphi consensus process, following recommended standards, was conducted. Participants (N=329) comprised three stakeholder groups (Researchers, n=58 (18%); Clinicians, n=247 (75%); Patients and Caregivers, n=24 (7%)), and represented 26 countries and 9 healthcare professions. Participants rated the importance of a range of relevant outcomes. Outcomes included in the COS were those prioritised of "critical importance" by all three stakeholder groups. Results Survey response rates were 88% (Round 1) and 91% (Round 2). From a total of 32 initial outcomes, the following outcomes reached consensus for inclusion in the COS: Physical Function, Activities of Daily Living, Survival, Health-related Quality of Life, Exercise Capacity, Cognitive Function, Emotional and Mental Wellbeing, and Frailty. Conclusion This study developed a consensus-generated COS for future clinical research evaluating physical rehabilitation interventions in critically ill adults across the continuum of recovery. Ascertaining recommended measurement instruments for these core outcomes is now required to facilitate implementation of the COS. This article is open access and distributed under the terms of the Creative Commons Attribution 4.0 International License (https://creativecommons.org/licenses/by/4.0/).