We examined whether unilateral exercise creates perception bias in the non-exercised limb and ascertained whether rTMS applied to the primary motor cortex (M1) interferes with this perception. All participants completed 4 interventions: 1) 15-min learning period of intermittent isometric contractions at 35% MVC with the trained hand (EX), 2) 15-min learning period of intermittent isometric contractions at 35% MVC with the trained hand whilst receiving rTMS over the contralateral M1 (rTMS+EX); 3) 15-min of rTMS over the 'trained' M1 (rTMS) and 4) 15-min rest (Rest). Pre and post-interventions, the error of force output production, the perception of effort (RPE), motor evoked potentials (MEPs) and compound muscle action potentials (CMAPs) were measured in both hands. EX did not alter the error of force output production in the trained hand (Δ3%; P>0.05); however, the error of force output production was reduced in the untrained hand (Δ12%; P<0.05). rTMS+EX and rTMS alone did not show an attenuation in the error of force output production in either hand. EX increased RPE in the trained hand (9.1±0.5 vs. 11.3±0.7; P<0.01) but not the untrained hand (8.8±0.6 vs. 9.2±0.6; P>0.05). RPE was significantly higher after rTMS+EX in the trained hand (9.2±0.5 vs. 10.7±0.7; P<0.01) but ratings were unchanged in the untrained hand (8.5±0.6 vs. 9.2±0.5; P>0.05). The novel finding was that exercise alone reduced the error in force output production by over a third in the untrained hand. Further, when exercise was combined with rTMS the transfer of force perception was attenuated. These data suggest that the contralateral M1 of the trained hand might, in part, play an essential role for the transfer of force perception to the untrained hand.