Abstract Background Early mobility therapy in the intensive care unit (ICU) is widely employed to improve the physical function and prognosis of patients with critically ill. On the other hand, patients who undergo cardiovascular surgery frequently suffer from disabilities after ICU care due to their cardiopulmonary conditions and treatments. However, few studies have reported the procedures and assessments of acute-phase rehabilitation in these patients. Recently, the Perme ICU Mobility Score (Perme Score) was developed as a reliable tool to assess comprehensive mobility status of patients in the ICU. We hypothesised that the Perme Score is a useful tool for assessing the mobility levels in the ICU and predicting clinical outcomes in patients undergoing acute-phase rehabilitation after cardiovascular surgery. Purpose To investigate the associations between the Perme Score within the second days after cardiovascular surgery and the patients' clinical outcomes, including physical function and the incidence of clinical events. Methods We studied 224 consecutive patients (34.4% female; aged 65±13 years) who were admitted to the ICU of a tertiary academic hospital after cardiac and/or major vascular surgery. Clinical characteristics including patient profiles, comorbidities, surgical details and APACHE II and SOFA scores were evaluated on ICU admission. The Perme Score contains categories on mental status, potential mobility barriers, muscle strength and mobility level, with higher scores indicating greater activity levels in the ICU. We assessed the Perme Score within the second days after the surgery. As a physical function at hospital discharge, we measured the six-minute walk distance (6MWD). The primary endpoint was a composite outcome of the number of all-cause mortality and/or all-cause unplanned readmission. We analysed the associations of the Perme Score with the 6MWD and the incidence of clinical events using multiple regression analysis and multivariate Poisson regression analysis, respectively. Results After adjusting for clinical confounding factors, a higher Perme Score was an independent factor of a higher 6MWD (Table 1). During the median follow-up period of 1.3 years, 51 cases of all-cause mortality/readmission occurred in 37 (16.5%) patients, with an incidence rate of 18.6/100 person-years. In the multivariate Poisson regression analysis, even after adjusting for the severity score in the ICU, a higher Perme Score was significantly and independently associated with lower rates of all-cause clinical events (adjusted incident rate ratio: 0.96, 95% confidence interval: 0.93–0.99, P=0.008, Figure 1). Conclusions The Perme Score within the second days after cardiovascular surgery is correlated with physical function at hospital discharge and the incidence of clinical events after discharge. Thus, a comprehensive assessment of acute-phase rehabilitation after cardiovascular surgery may be useful in predicting clinical outcomes. Funding Acknowledgement Type of funding sources: None.