Whilst arm cycling is commonly used during clinical testing and rehabilitation, the associated changes in arterial plasma K+ concentration ([K+]a) and thus possible risks of associated hyperkalaemia and hypokalaemia are unknown. During exercise, K+ is released from contracting skeletal muscles and simultaneously taken up by inactive muscles, with [K+]a increasing markedly during intense exercise. Hence during exercise [K+]a should be influenced by both the contracting muscle mass and the relative exercise intensity. We therefore investigated the effects of incremental one-arm (1ARM) and two-leg (2LEGS) cycling on [K+]a, and on lactate ([Lac-]), hydrogen ([H+]), adrenaline and noradrenaline. Eight healthy participants performed 1ARM, rested (3.0 ± 0.2 h, mean±SE) then performed 2LEGS, with radial arterial blood sampled at rest, during common submaximal workrates (25, 50 W), peak workrate and 1–30 min post-exercise. During exercise at 50 W, VO2, [K+]a and [Lac-]a were higher during 1ARM than 2LEGS (P < 0.05). In contrast, at peak exercise, 2LEGS (272 ± 15 W) elicited higher VO2, [K+]a, [Lac-]a, [adrenaline]a and [noradrenaline]a than 1ARM (57 ± 4 W) (P < 0.05), with [H+]a elevated only in 2LEGS (P < 0.05). A curvilinear response for [K+]a versus %VO2peak was similar between modes, until diverging close to VO2peak. The slope of the log[K+]a versus %VO2peak regression was higher in 2LEGS (P < 0.01), indicating greater [K+]a throughout exercise utilising a larger muscle mass, and similarly for [Lac-]a. In conclusion, during exercise utilising a smaller contracting muscle mass, [K+]a and [Lac-]a were greater at the same absolute submaximal work rate compared to utilising a large muscle mass. Whilst responses were similar between modes when expressed against relative exercise intensity, [K+]a and [Lac-]a were higher during exercise with a large muscle mass. Hence, for submaximal exercise, the relative intensity is more important in determining [K+]a and [Lac-]a, but the size of the contracting muscle mass still exerts a positive effect, whilst during peak exercise, the size of the active muscle mass was more important in determining [K+]a and [Lac-]a. Furthermore, arm cycling did not induce marked disturbances in [K+]a during or after exercise, suggesting minimal associated myocardial risks and thus supporting its safe use in clinical and aged populations.