Cervical or upper-thoracic spinal cord injury (SCI, ≥T6) often leads to low resting blood pressure (BP) and impaired cardiovascular responses to acute exercise due to disrupted supraspinal sympathetic drive. Epidural spinal cord stimulation (invasive, ESCS) and transcutaneous spinal cord stimulation (non-invasive, TSCS) have previously been used to target dormant sympathetic circuits and modulate cardiovascular responses. This case series compared the effects of cardiovascular-optimised ESCS and TSCS versus sham ESCS and TSCS on modulating cardiovascular responses and improving submaximal upper-body exercise performance in individuals with SCI. Seven males with a chronic, motor-complete SCI between C6 and T4 underwent a mapping session to identify cardiovascular responses to spinal cord stimulation. Subsequently, four participants (two ESCS and two TSCS) completed submaximal exercise testing. Stimulation parameters (waveform, frequency, intensity, epidural electrode array configuration, and transcutaneous electrode locations in the lumbosacral region) were optimised to elevate cardiovascular responses (CV-SCS). A sham condition (SHAM-SCS) served as a comparison. Participants performed arm-crank exercise to exhaustion at a fixed workload corresponding to above ventilatory threshold, on separate days, with CV-SCS or SHAM-SCS. At rest, CV-SCS increased BP and predicted left ventricular cardiac contractility and total peripheral resistance. During exercise, CV-SCS increased time to exhaustion and peak oxygen pulse (a surrogate for stroke volume), relative to SHAM-SCS. Ratings of perceived exertion also tended to be lower with CV-SCS than SHAM-SCS. Comparable improvements in time to exhaustion with ESCS and TSCS suggest that both approaches could be promising ergogenic aids to support exercise performance or rehabilitation, along with reducing fatigue during activities of daily living in individuals with SCI.