The present study investigated the effects of exercise on the serum concentrations of brain-derived neurotrophic factor (BDNF), insulin-like growth factor 1 (IGF-1), prolactin (PRL) and cortisol (COR) in 11 chronically spinal cord-injured athletes. In these subjects BDNF concentration at rest was sixfold higher compared with the concentrations reported earlier in able-bodied persons, while IGF-1, PRL and COR were within normal range. Ten minutes of moderate intensity handbiking (54% of the maximal heart rate) during a warm-up period (W) induced an increase (P<0.05) of BDNF of approximately 1.5-fold from basal level at rest, while a decrease to basal level was found after an immediately succeeding handbiking time trial (89% of the maximal heart rate) over the marathon distance of 42 km (M). An increase (P<0.01) of serum IGF-1 was found after W and this levels remained elevated (P<0.01) until the end of M. W had no significant effects on the serum PRL and COR, however, M induced an increase (P<0.01) of both hormones. This is the first study showing elevated BDNF concentrations at rest in spinal cord-injured athletes. Furthermore, short moderate intensity handbiking but not immediately following long lasting high intensity handbiking further increases serum BDNF concentrations. IGF-1 response to exercise differs to BDNF response as this neuroplasticity-related protein remains elevated during the long lasting physical demand with high intensity. The augmented PRL concentration suggests that a possible mechanism by which exercise promotes neuroplasticity might be the activation of neural serotonergic pathways as 5-HT is the main PRL releasing factor. Elevated COR concentrations after M are unlikely to be deleterious to neuroplasticity as COR concentrations remain within the physiological range. The present study suggests that exercise might be beneficial to enhance neuroprotection and neuroplasticity, thereby improving recovery after spinal cord injury.