Twelve adolescent athletes underwent, in a crossover-design study, 3 separate 90-min training sessions in the following conditions: no fluid ingestion allowed (NF), ad libitum ingestion of water (W), and ad libitum ingestion of a commercial 8% carbohydrate-electrolyte sports beverage (CSB). After each session athletes performed a set of basketball drills (2-point, 3-point, and free-throw shootout, suicide sprints, and defensive zigzags). Body weight (before and after sessions), rating of perceived exertion (RPE), urine color, and beverage acceptability were determined in each session. Athletes also completed a survey about their knowledge and behaviors regarding hydration and fluid replacement. The percentage of weight loss was significantly higher in NF (2.46% ± 0.87%) than in the other 2 conditions (W, 1.08% ± 0.67%, p = .006; CSB, 0.65% ± 0.62%, p = .001) but also higher in W than CSB (p = .012). RPE was higher in NF (16.8 ± 1.96) than in the W (14.2 ± 1.99, p = .004) and CSB (13.3 ± 2.06, p = .002) trials. Athletes' fluid intake was positively correlated with proper self-reported behaviors (r = .75, p = .005) and knowledge (r = .76, p = .004) about fluid and hydration. In conclusion, fluid restriction during exercise was associated with a greater level of dehydration and increased perceived exertion but had no impact on basketball performance compared with ad libitum drinking of water or a CSB. Athletes with more knowledge about hydration and better self-reported hydration behaviors ingested more fluids during training sessions.