PURPOSE: Night duty has been recognized as a significantly harmful stressor for physicians. However, the relationship between various levels of duty loading and stress response is unknown. This study examined whether duty load increases cardiovascular stress indicators in a dose-dependent manner. METHODS: An unallocated prospective observational study was conducted among physicians performing various levels of duties in a secondary referral medical center between 2011 and 2012. Heart rate variability (HRV), blood pressure (BP), and other stress markers of 12 attending physicians were compared during different duty loads: non-duty day (NDD), duty day with one duty area and three wards (1DD), and duty day with two duty areas and six wards (2DD). RESULTS: During the regular sleep time (i.e., 11 p.m. to 5 a.m.), the relative sympathetic modulations measured using the HRV were 59.0 +/- 9.3, 61.6 +/- 10.4, and 64.4 +/- 8.9 for NDD, 1DD, and 2DD, respectively (p = 0.0012); those for relative parasympathetic modulations were 37.4 +/- 9.4, 34.8 +/- 9.8, and 32.0 +/- 8.8 for NDD, 1DD, and 2DD, respectively (p = 0.0015). The percentages of abnormal systolic BPs were 9.7 +/- 13.2 %, 25.3 +/- 21.8 %, and 31.5 +/- 21.0 % for NDD, 1DD, and 2DD, respectively (p = 0.003), and the percentages of abnormal diastolic BP were 6.7 +/- 11.0 %, 18.3 +/- 11.1 %, and 27.1 +/- 30.9 % for NDD, 1DD, and 2DD, respectively (p = 0.002). Total sleep time was negatively associated with sympathetic/parasympathetic balance and the percentage of abnormal diastolic BP. Admitting new patients was positively associated with the percentages of abnormal systolic BP. CONCLUSIONS: This observational analysis suggests that the dose-dependent stress responses of the cardiovascular system in physicians were caused by the duty load.