PURPOSE: Cold water immersion (CWI) may be beneficial for acute recovery from exercise, but it may impair long-term performance by attenuating the stimuli responsible for adaptation to training. We compared effects of CWI and passive rest on cycling performance during a simulated cycling grand tour. METHODS: Thirty-four male endurance-trained competitive cyclists were randomized to CWI for four times per week for 15 min at 15°C or control (passive recovery) groups for 7 d of baseline training, 21 d of intensified training, and an 11-d taper. Criteria for completion of training and testing were satisfied by 10 cyclists in the CWI group (maximal aerobic power, 5.13 ± 0.21 W·kg; mean ± SD) and 11 in the control group (5.01 ± 0.41 W·kg). Each week, cyclists completed a high-intensity interval cycling test and two 4-min bouts separated by 30 min. CWI was performed four times per week for 15 min at 15°C. RESULTS: Between baseline and taper, cyclists in the CWI group had an unclear change in overall 4-min power relative to control (2.7% ± 5.7%), although mean power in the second effort relative to the first was likely higher for the CWI group relative to control (3.0% ± 3.8%). The change in 1-s maximum mean sprint power in the CWI group was likely beneficial compared with control (4.4% ± 4.2%). Differences between groups for the 10-min time trial were unclear (-0.4% ± 4.3%). CONCLUSION: Although some effects of CWI on performance were unclear, data from this study do not support recent speculation that CWI is detrimental to performance after increased training load in competitive cyclists.