The purpose of this study was to assess the influence of aerobic training on the left ventricular (LV) systolic function. Thirty patients with stable coronary artery disease, who had participated in the conducted 3-month physical training, were retrospectively divided into 2 cohorts. While patients in the cohort I (n=14) had continued training individually for 12 months, patients in the cohort II (n=16) had stopped training after finishing the conducted program. Rest and stress dobutamine/atropine echocardiography was performed in all patients before the training program and 1 year later. The peak systolic velocities of mitral annulus (Sa) were assessed by tissue Doppler imaging for individual LV walls. In addition, to determine global LV systolic longitudinal function, the four-site mean systolic velocity was calculated (Sa glob). According to the blood supply, left ventricular walls were divided into 5 groups: A- walls supplied by nonstenotic artery; B- walls supplied by coronary artery with stenosis ≤50 %; C- walls supplied by coronary artery with stenosis 51-70 %; D- walls with stenosis of supplying artery 71-99 %; and E- walls with totally occluded supplying artery. In global systolic function, the follow-up values of Sa glob in cohort I were improved by 0.23±0.36 as compared with baseline values at rest, and by 1.26±0.65 cm/s at the maximal load, while the values of Sa glob in cohort II were diminished by 0.53±0.22 (p=NS), and by 1.25±0.45 cm/s (p<0.05), respectively. Concerning the resting regional function, the only significant difference between cohorts in follow-up changes was found in walls E: 0.37±0.60 versus -1.76±0.40 cm/s (p<0.05). At the maximal load, the significant difference was found only in walls A (0.16±0.84 versus -2.67±0.87 cm/s; p<0.05). Patients with regular 12-month physical activity improved their global left ventricle systolic function mainly due to improvement of contractility in walls supplied by a totally occluded coronary artery.