Objective: To evaluate the feasibility and safety of exercise testing and to describe the physiological response to exercise of patients in the Intensive Care Unit (ICU). Design: A prospective observational multicenter study. Setting: Two mixed medical-surgical ICUs. Participants: Patients (N=37; with no primary neurological disorders, 59% men; median age 50y; ICU length of stay 14.5d; Acute Physiology and Chronic Health Evaluation IV 73.0) who had been mechanically ventilated for more than 48 hours and were hemodynamically stable enough to perform physical exercise. Interventions: A passive or active incremental exercise test, depending on muscle strength, on a bed-based cycle ergometer. Main Outcome Measures: Feasibility and safety were evaluated based on protocol adherence and adverse events. Physiological responses to exercise quantified as changes in respiratory frequency (RF), oxygen uptake (VO2), carbon dioxide output (VCO2), respiratory exchange ratio (RER), and blood lactate. Results: Thirty-seven patients of whom 18 were mechanically ventilated underwent the exercise test. The active incremental test was performed by 28, and the passive test by 9 participants. Thirty-three (89%) accomplished the test according to the protocol and 1 moderate severe adverse event (bradycardia; heart rate 44) occurred shortly after the test. RF, VO2, VCO2, and lactate increased significantly, whereas RER did not change during the active incremental exercise test. No changes were observed during the passive exercise test. Conclusions: It is safe and feasible to perform exercise testing on a bed-based cycle ergometer in patients who are critically ill and a physiological response could be measured. Future research should investigate the clinical value of exercise testing in daily ICU practice and whether exercise capacity and its limiting factors could be determined by incremental exercise testing.