Background and purpose Much is known about cardiovascular and biomechanical responses to exercise during water immersion, yet an understanding of the higher-order neural responses to water immersion is unclear. The purpose of this study was to compare cognitive and motor performance between land and water environments using a dual-task paradigm, which served as an indirect measure of cortical processing. Design A quasi-experimental crossover research design is used. Methods Twenty-two healthy participants (age = 24.3 ± 5.24 years) and a single-case patient (age = 73) with mild cognitive impairment performed a cognitive (auditory vigilance) and motor (standing balance) task separately (single-task condition) and simultaneously (dual-task condition) on land and in chest-deep water. Listening errors from the auditory vigilance task and centre of pressure (CoP) area for the balance task measured cognitive and motor performance, respectively. Results Listening errors for the single-task and dual-task conditions were 42% and 45% lower for the water than land condition, respectively (effect size [ES] = 0.38 and 0.55). CoP area for the single-task and dual-task conditions, however, were 115% and 164% lower on land than in water, respectively, and were lower (≈8–33%) when balancing concurrently with the auditory vigilance task compared with balancing alone, regardless of environment (ES = 0.23–1.7). This trend was consistent for the single-case patient. Conclusion Participants tended to make fewer ‘cognitive’ errors while immersed chest-deep in water than on land. These same participants also tended to display less postural sway under dual-task conditions, but more in water than on land. Copyright © 2015 John Wiley & Sons, Ltd.