The aim of this study was to test the hypothesis that adding caffeine to postexercise carbohydrate (CHO) feedings improves subsequent high-intensity interval-running capacity compared with CHO alone. In a repeated-measures design, 6 men performed a glycogen-depleting exercise protocol until volitional exhaustion in the morning. Immediately after and at 1, 2, and 3 hr postexercise, participants consumed 1.2 g/kg body mass CHO of a 15% CHO solution, a similar CHO solution but with addition of 8 mg/kg body mass of caffeine (CHO+CAFF), or an equivalent volume of flavored water only (WAT). After the 4-hr recovery period, participants performed the Loughborough Intermittent Shuttle Test (LIST) to volitional exhaustion as a measure of high-intensity interval-running capacity. Average blood glucose values during the 4-hr recovery period were higher in the CHO conditions (p < .005) than in the WAT trial (4.6 ± 0.3 mmol/L), although there was no difference (p = .46) between CHO (6.2 ± 0.8 mmol/L) and CHO+CAFF (6.7 ± 1.0 mmol/L). Exercise capacity during the LIST was significantly longer in the CHO+CAFF trial (48 ± 15 min) than in the CHO (32 ± 15 min, p = .04) and WAT conditions (19 ± 6 min, p = .001). All 6 participants improved performance in CHO+CAFF compared with CHO (95% CI for mean difference = 1-32 min). The study provides novel data by demonstrating that adding caffeine to postexercise CHO feeding improves subsequent high-intensity interval-running capacity, a finding that may be related to higher rates of postexercise muscle glycogen resynthesis previously observed under similar feeding conditions.