The purpose of this study was to compare the EMG activity of blood flow restricted (limb) and nonrestricted (trunk) muscles during multi-joint exercise with and without KAATSU. Twelve (6 women and 6 men) healthy college students [means (SD) age: 24.1 (3.5) yrs] performed 4 sets (30, 15, 15, and 15 reps) of flat bench press exercise (30% of a predetermined one repetition maximum, 1-RM) during two different conditions [with KAATSU and without KAATSU (Control)]. In the KAATSU condition, a specially designed elastic cuff belt (30 mm wide) was placed at the most proximal position of the upper arm and inflated to a pressure of 100% of individual's resting systolic blood pressure. Surface EMG was recorded from the muscle belly of the triceps brachii (TB) and pectoralis major (PM) muscles, and mean integrated EMG (iEMG) was analyzed. During 4 sets of the exercise, gradual increases in iEMG were observed in both TB and PM muscles for the KAATSU condition. The magnitude of the increases in iEMG in the TB and PM muscles were higher (P<0.05) with KAATSU compared to the Control condition. In the first set, the mean exercise intensity from normalized iEMG was approximately 40% of 1-RM in both Control and KAATSU conditions. However, the mean exercise intensity of both muscles were 60-70% of 1-RM for the KAATSU condition and only about 50% of 1-RM for the Control condition, respectively, during the fourth set. We concluded that increases in iEMG in the trunk muscle during KAATSU might be an important factor for KAATSU training-induced trunk muscle hypertrophy.