PURPOSE: To examine the effects of a high-carbohydrate diet (HCHO), periodized-carbohydrate (CHO) diet (PCHO), and ketogenic low-CHO high-fat diet (LCHF) on training capacity. METHODS: Elite male racewalkers completed 3 weeks of periodic training while adhering to their dietary intervention. Twenty-nine data sets were collected from 21 athletes. Each week, 6 mandatory training sessions were completed, with additional sessions performed at the athlete's discretion. Mandatory sessions included an interval session (10 × 1-km efforts on a 6-min cycle), tempo session (14 km with a 450-m elevation gain), 2 long walks (25-40 km), and 2 easy walks (8-12 km) where "sleep-low" and "train-low" dietary strategies were employed for PCHO. Racewalking speed, heart rate, rating of perceived exhaustion, and blood metabolites were collected around key sessions. RESULTS: LCHF covered less total distance than HCHO and PCHO (P < .001); however, no differences in training load between groups were evident (P = .285). During the interval sessions, walking speed was slower in LCHF (P = .001), equating to a 2.8% and 5.6% faster speed in HCHO and PCHO, respectively. LCHF was also 3.2% slower in completing the tempo session than HCHO and PCHO (P = .001). Heart rate was higher (P = .002) and lactate concentrations were lower (P < .001) in LCHF compared to other groups, despite slower walking speeds during the interval session. No between-groups differences in rating of perceived exhaustion were evident (P = .077). CONCLUSION: Athletes adhering to an LCHF diet showed impaired training capacity relative to their high-CHO-supported counterparts, completing lower training volumes at slower speeds, with higher heart rates.