Key points Blood glucose is an important fuel for endurance exercise. It can be derived from ingested carbohydrate, stored liver glycogen and newly synthesized glucose (gluconeogenesis). We hypothesized that athletes habitually following a low carbohydrate high fat (LCHF) diet would have higher rates of gluconeogenesis during exercise compared to those who follow a mixed macronutrient diet. We used stable isotope tracers to study glucose production kinetics during a 2 h ride in cyclists habituated to either a LCHF or mixed macronutrient diet. The LCHF cyclists had lower rates of total glucose production and hepatic glycogenolysis but similar rates of gluconeogenesis compared to those on the mixed diet. The LCHF cyclists did not compensate for reduced dietary carbohydrate availability by increasing glucose synthesis during exercise but rather adapted by altering whole body substrate utilization. Abstract Endogenous glucose production (EGP) occurs via hepatic glycogenolysis (GLY) and gluconeogenesis (GNG) and plays an important role in maintaining euglycaemia. Rates of GLY and GNG increase during exercise in athletes following a mixed macronutrient diet; however, these processes have not been investigated in athletes following a low carbohydrate high fat (LCHF) diet. Therefore, we studied seven well‐trained male cyclists that were habituated to either a LCHF (7% carbohydrate, 72% fat, 21% protein) or a mixed diet (51% carbohydrate, 33% fat, 16% protein) for longer than 8 months. After an overnight fast, participants performed a 2 h laboratory ride at 72% of maximal oxygen consumption. Glucose kinetics were measured at rest and during the final 30 min of exercise by infusion of [6,6‐2H2]‐glucose and the ingestion of 2H2O tracers. Rates of EGP and GLY both at rest and during exercise were significantly lower in the LCHF group than the mixed diet group (Exercise EGP: LCHF, 6.0 ± 0.9 mg kg−1 min−1, Mixed, 7.8 ± 1.1 mg kg−1 min−1, P < 0.01; Exercise GLY: LCHF, 3.2 ± 0.7 mg kg−1 min−1, Mixed, 5.3 ± 0.9 mg kg−1 min−1, P < 0.01). Conversely, no difference was detected in rates of GNG between groups at rest or during exercise (Exercise: LCHF, 2.8 ± 0.4 mg kg−1 min−1, Mixed, 2.5 ± 0.3 mg kg−1 min−1, P = 0.15). We conclude that athletes on a LCHF diet do not compensate for reduced glucose availability via higher rates of glucose synthesis compared to athletes on a mixed diet. Instead, GNG remains relatively stable, whereas glucose oxidation and GLY are influenced by dietary factors.