Background Perceived exertion is an inevitable outcome of power-oriented resistance training. However, it remains unknown whether perceived exertion is dominated by central or peripheral responses during this type of training. This study examined the effect of neuromuscular responses on the determination of ratings of perceived exertion (RPE) during power bench press (BPress) exercise. Methods Sixteen trained males performed three BPress tasks with varying volumes (low, medium, high) at 65% of their one-repetition maximum. RPE, surface electromyography, and velocity loss were assessed across all conditions. Peak root mean square (RMS) and median frequency (MDF) were calculated from the surface electromyography data. Results Significant effects were observed across experimental conditions for overall RPE, average velocity loss, and average MDF (all p < 0.001), while no significant difference was found in average RMS. As the lifting tasks progressed, significant effects of repetition were observed in all measured variables (all p < 0.001). When comparing conditions, significant differences were found among the three in RPE, velocity loss, and MDF (all p < 0.001), whereas no significant effect of condition was observed in RMS. No significant relationship was found between MDF and velocity loss. Conclusions In power BPress, higher repetitions affected RPE, velocity loss, and MDF, while peak RMS was less responsive. These findings indicate that both athletic performance and RPE are primarily influenced by peripheral fatigue. However, velocity loss should be interpreted with caution as a fatigue marker in this context.