Very little is known about how long-term (>6 months) adaptation to a low-carbohydrate, high-fat (LCHF) diet affects insulin signaling in healthy, well-trained individuals. This study compared glucose tolerance; skeletal muscle glucose transporter 4 (GLUT4) and insulin receptor substrate 1 (IRS1) content; and muscle enzyme activities representative of the main energy pathways (3-hydroxyacetyl-CoA dehydrogenase, creatine kinase, citrate synthase, lactate dehydrogenase, phosphofructokinase, phosphorylase) in trained cyclists who followed either a long-term LCHF or a mixed-macronutrient (Mixed) diet. On separate days, a 2-hr oral glucose tolerance test was conducted, and muscle samples were obtained from the vastus lateralis of fasted participants. The LCHF group had reduced glucose tolerance compared with the Mixed group, as plasma glucose concentrations were significantly higher throughout the oral glucose tolerance test and serum insulin concentrations peaked later (LCHF, 60 min; Mixed, 30 min). Whole-body insulin sensitivity was not statistically significantly different between groups (Matsuda index: LCHF, 8.7 ± 3.4 vs. Mixed, 12.9 ± 4.6; p = .08). GLUT4 (LCHF: 1.13 ± 0.24; Mixed: 1.44 ± 0.16; p = .026) and IRS1 (LCHF: 0.25 ± 0.13; Mixed: 0.46 ± 0.09; p = .016) protein content was lower in LCHF muscle, but enzyme activities were not different. We conclude that well-trained cyclists habituated to an LCHF diet had reduced glucose tolerance compared with matched controls on a mixed diet. Lower skeletal muscle GLUT4 and IRS1 contents may partially explain this finding. This could possibly reflect an adaptation to reduced habitual glucose availability rather than the development of a pathological insulin resistance.