Purpose: To examine the effects of daily cold- and hot-water recovery on training load (TL) during 5 days of heat-based training. Methods: Eight men completed 5 days of cycle training for 60 minutes (50% peak power output) in 4 different conditions in a block counter-balanced-order design. Three conditions were completed in the heat (35°C) and 1 in a thermoneutral environment (24°C; CON). Each day after cycling, participants completed 20 minutes of seated rest (CON and heat training [HT]) or cold- (14°C; HTCWI) or hot-water (39°C; HTHWI) immersion. Heart rate, rectal temperature, and rating of perceived exertion (RPE) were collected during cycling. Session-RPE was collected 10 minutes after recovery for the determination of session-RPE TL. Data were analyzed using hierarchical regression in a Bayesian framework; Cohen d was calculated, and for session-RPE TL, the probability that d > 0.5 was also computed. Results: There was evidence that session-RPE TL was increased in HTCWI (d = 2.90) and HTHWI (d = 2.38) compared with HT. The probabilities that d > 0.5 were .99 and .96, respectively. The higher session-RPE TL observed in HTCWI coincided with a greater cardiovascular (d = 2.29) and thermoregulatory (d = 2.68) response during cycling than in HT. This result was not observed for HTHWI. Conclusion: These findings suggest that cold-water recovery may negatively affect TL during 5 days of heat-based training, hot-water recovery could increase session-RPE TL, and the session-RPE method can detect environmental temperature-mediated increases in TL in the context of this study.