Purpose The present study investigated the effects of periodic carbohydrate (CHO) restriction on endurance performance and metabolic markers in elite endurance athletes. Methods Twenty-six male elite endurance athletes (maximal oxygen consumption (VO 2max), 65.0 mL O 2·kg -1·min -1) completed 4 wk of regular endurance training while being matched and randomized into two groups training with (low) or without (high) CHO manipulation 3 d·wk -1. The CHO manipulation days consisted of a 1-h high-intensity bike session in the morning, recovery for 7 h while consuming isocaloric diets containing either high CHO (414 ± 2.4 g) or low CHO (79.5 ± 1.0 g), and a 2-h moderate bike session in the afternoon with or without CHO. VO 2max, maximal fat oxidation, and power output during a 30-min time trial (TT) were determined before and after the training period. The TT was undertaken after 90 min of intermittent exercise with CHO provision before the training period and both CHO and placebo after the training period. Muscle biopsies were analyzed for glycogen, citrate synthase (CS) and β-hydroxyacyl-coenzyme A dehydrogenase (HAD) activity, carnitine palmitoyltransferase (CPT1b), and phosphorylated acetyl-CoA carboxylase (pACC). Results The training effects were similar in both groups for all parameters. On average, VO 2max and power output during the 30-min TT increased by 5% ± 1% (P < 0.05) and TT performance was similar after CHO and placebo during the preload phase. Training promoted overall increases in glycogen content (18% ± 5%), CS activity (11% ± 5%), and pACC (38% ± 19%; P < 0.05) with no differences between groups. HAD activity and CPT1b protein content remained unchanged. Conclusions Superimposing periodic CHO restriction to 4 wk of regular endurance training had no superior effects on performance and muscle adaptations in elite endurance athletes.