Cancer patients lose weight due to negative energy balance because of insufficient appetite and inappropriately high energy expenditure. Host and tumor derived cytokines and more recently eicosanoids have been held responsible as mediators. Accordingly, observations in animal experiments and short-term clinical trials in selected groups of cancer patients, have implied that cyclo-oxygenase (COX) blockade can improve host metabolism and well-being, and long-term COX-treatment of unselected groups have implied improved survival. The aim of this study was to search for evidence that long-term COX-treatment improves energy and cardiovascular homeostasis in unselected weight-losing cancer patients. A retrospective case control analysis was performed on a data-base material collected consecutively. Weight-losing untreated cancer patients had elevated resting energy expenditure compared to undernourished non-cancer patients (23.3+/-0.1, n=702 vs 20.9+/-0.3 kcal/kg/day, n=132, p<0.001). This difference became significantly reduced by long-term indomethacin treatment (p<0.003). Heart rate was correspondingly decreased, while systolic blood pressure increased following indomethacin treatment of cancer patients (p<0.006-0.008). Total body fat was more preserved (p<0.005), while lean body mass was uninfluenced by long-term indomethacin to cancer patients. All these beneficial effects were parallel to a decrease in systemic inflammation (C-reactive protein, erythrocyte sedimentation rate) in cancer patients on indomethacin (p<0.0004). Systemic inflammation and resting energy metabolism predicted weight loss in progressive cancer (p<0.0001). Our data support the concept that COX-treatment may offer beneficial metabolic effects to weight-losing cancer patients by attenuation of resting metabolism and improved appetite due to decreased systemic inflammation.