Abstract Prehabilitation is being widely adopted to optimise physical health prior to surgery. However, there remains a lack of consensus regarding exercise type and intensity with current guidance extrapolated from generic, non-cancer populations. The WHO and American College of Sports Medicine (ACSM) recommend 600-1200 Metabolic Equivalent minutes (MET-mins) per week of physical activity (PA) but the effectiveness of this recommendation in cancer populations is unknown. Patients undergoing treatment for Oesophagogastric cancer were prospectively recruited. International Physical Activity Questionnaire (IPAQ) quantified PA at baseline and throughout treatment. Two cohorts were defined (lowPA and highPA) based on ACSM cut-off of 1200 MET-mins/week. Outcome measures were physical function (1 minute sit-to-stand, 1mSTS), CT muscle mass, chemotoxicity, and failure to complete planned treatment. 38 patients completed all study assessments (highPA n=20, lowPA n=18) with no difference in baseline characteristics or disease stage between groups. LowPA group showed declines in 1mSTS performance following chemotherapy, whereas the highPA group improved (mean difference -4 versus +10, p=0.001). Whilst both saw a decline in skeletal muscle mass, the lowPA group lost significantly more (mean difference -5.1 versus -1.8, p=0.007). There were trends towards a lower incidence of severe chemotoxicity (30% vs 61%, p=0.054) and failure to complete planned treatment (15% versus 28%, p=.33) in the highPA group. Whilst further research is required to define the most effective prehabilitation strategy, this study is the first to suggest a new benchmark to preserve muscle mass and function is higher than current guidelines and should be considered when developing prehabilitation recommendations.