The study compares power outputs in the concentric phase of chest presses and squats performed in the interval mode on stable and unstable surface, respectively. A group of 16 physical education students performed randomly on different days 6 sets of 8 repetitions of (a) chest presses on the bench and Swiss ball, respectively, and (b) squats on stable support base and Bosu ball, respectively, with 2 minutes of rest period between sets. The exercises were performed with previously established 70% of 1 repetition maximum under stable conditions. A PC-based system FiTRO Dyne Premium was used to monitor force and velocity and to calculate power. The results showed significantly lower power outputs when resistance exercises were performed on an unstable than a stable support base. In the initial set, mean power in concentric phase of lifting decreased more profoundly under unstable than under stable conditions during both chest presses (13.2 and 7.7%, respectively) and squats (10.3 and 7.2%, respectively). In the final set, the reduction rates of mean power in the concentric phase of chest presses were significantly (p < 0.05) greater on the Swiss ball than on the bench (19.9 and 11.8%, respectively). On the other hand, there were no significant differences in decline of mean power in the concentric phase of squats on the Bosu ball and on stable support base (11.4 and 9.6%, respectively). It may be concluded that power outputs during resistance exercises is more profoundly compromised under unstable than under stable conditions, and this effect is more evident for barbell chest presses on the Swiss ball than for barbell squats on the Bosu ball. These findings have to be taken into account when instability resistance exercises are implemented into the training program, namely, for sports that require production of maximal force in short time.