Training with low carbohydrate availability enhances endurance training adaptations but training volume may be compromised. We explored whole-body metabolism and performance with delayed carbohydrate feeding during exercise undertaken following acute sleep-low training. We hypothesised this strategy would not suppress fat oxidation and would maintain exercise performance. The study involved three experimental trials and included 9 men and 1 woman (⩒O2peak = 58.8 ± 5.5 mL kg−1 min−1). Each trial started in the afternoon with an exhaustive cycling protocol. The following morning 1-h of steady-state cycling (SS) was followed by a time trial (TT). Carbohydrates (CHO) were not ingested in recovery from exhaustive exercise or during next day exercise in the Placebo trial (PLA); CHO were not ingested during recovery but were fed (15 g every ∼15-min) from 30-min into SS and continued during the TT in the delayed feeding trial (DELAY); CHO were provided during recovery (1.2 g/kg/h for 7 h) and next day exercise (as in DELAY) in a third condition (CHO). Exercise metabolism was assessed using indirect calorimetry and blood sampling. Fat oxidation rates during SS were similar in PLA (0.83 ± 0.17 g/min) and DELAY (0.78 ± 0.14 g/min) (p > 0.05) and higher than CHO (0.57 ± 0.27 g/min) (p < 0.05). There were no significant differences in TT performance (49.1 ± 10.7, 43.4 ± 7.6, 41.0 ± 7.9 min in PLA, DELAY and CHO, respectively; p > 0.05). Delayed carbohydrate feeding could be a strategy to maintain high-fat oxidation rates typically associated with exercise undertaken after the sleep-low approach to training but the acute performance effects remain inconclusive.