Carbohydrate and fat are the main substrates utilized during prolonged endurance-type exercise. The relative contribution of each is determined primarily by the intensity and duration of exercise, along with individual training and nutritional status. During moderate- to high-intensity exercise, carbohydrate represents the main substrate source. Because endogenous carbohydrate stores (primarily in liver and muscle) are relatively small, endurance-type exercise performance/capacity is often limited by endogenous carbohydrate availability. Much exercise metabolism research to date has focused on muscle glycogen utilization, with little attention paid to the contribution of liver glycogen. (13)C magnetic resonance spectroscopy permits direct, noninvasive measurements of liver glycogen content and has increased understanding of the relevance of liver glycogen during exercise. In contrast to muscle, endurance-trained athletes do not exhibit elevated basal liver glycogen concentrations. However, there is evidence that liver glycogenolysis may be lower in endurance-trained athletes compared with untrained controls during moderate- to high-intensity exercise. Therefore, liver glycogen sparing in an endurance-trained state may account partly for training-induced performance/capacity adaptations during prolonged (>90 min) exercise. Ingestion of carbohydrate at a relatively high rate (>1.5 g/min) can prevent liver glycogen depletion during moderate-intensity exercise independent of the type of carbohydrate (e.g., glucose vs. sucrose) ingested. To minimize gastrointestinal discomfort, it is recommended to ingest specific combinations or types of carbohydrates (glucose plus fructose and/or sucrose). By coingesting glucose with either galactose or fructose, postexercise liver glycogen repletion rates can be doubled. There are currently no guidelines for carbohydrate ingestion to maximize liver glycogen repletion.