PURPOSE: To examine changes in brain derived neurotropic factor (BDNF) concentration in both plasma and serum following exercise in the cold. METHODS: Six recreationally active men (26±3 yrs; 180.3±5.8 cm; 85.3±8.4 kg; 48.6±5.7 ml·kg−1·min−1) completed an exercise protocol under two conditions: moderate temperature (MT; 23°C/45%RH) and low temperature (LT; 5°C). The protocol consisted of a 60-minute cycling trial at 60% VO2max, a 15-minute rest, and a time-to-exhaustion trial at 90% VO2max (TTE). Blood was collected before (T1) and after (T2) the 60-minute trial, immediately after TTE (T3), and one hour post-TTE. Plasma and serum concentrations of BDNF were measured via ELISA. Changes were analyzed using separate condition by time mixed-model regression for each dependent variable. RESULTS: No significant condition × time interaction (F =1.626, p =0.201) or main effect of time (F =0.626, p=0.603) was observed for changes in serum BDNF concentrations; however, a significant main effect of condition (F=7.685, p = 0.009) was observed. When collapsed across time, serum BDNF concentration was lower during LT (2718.8 1172.2 pg/mL) compared to MT (7240.5 1134.2 pg/mL; p =0.009). No significant condition × time interaction (F =0.117, p =0.950), main effect of time (F =0.511, p =0.677) nor main effect of condition (F =0.000, p =0.988) was observed for changes in plasma BDNF concentrations. CONCLUSIONS: The results of this study suggest that exercise in a cold environment blunts serum BDNF concentration. However, plasma concentrations of BDNF were not affected by environmental condition nor exercise.