Carbohydrate counting is a meal planning approach used with clients who have diabetes that focuses on carbohydrate as the primary nutrient affecting postprandial glycemic response. The concept of carbohydrate counting has been around since the 1920s, but it received renewed interest after being used as 1 of 4 meal planning approaches in the Diabetes Control and Complications Trial. In the trial, carbohydrate counting was found to be effective in meeting outcome goals and allowed flexibility in food choices. Recent practice pattern surveys have shown an increasing interest in and use of carbohydrate counting for medical nutrition therapy for persons with diabetes. Carbohydrate counting can be used by clients with type 1, type 2, and gestational diabetes. Three levels of carbohydrate counting have been identified based on increasing levels of complexity. Level 1, or basic, introduces clients to the concept of carbohydrate counting and focuses on carbohydrate consistency. Level 2, or intermediate, focuses on the relationships among food, diabetes medications, physical activity, and blood glucose level and introduces the steps needed to manage these variables based on patterns of blood glucose levels. Level 3, or advanced, is designed to teach clients with type 1 diabetes who are using multiple daily injections or insulin infusion pumps how to match short-acting insulin to carbohydrate using carbohydrate-to-insulin ratios. All 3 levels emphasize portion control and offer opportunities for using creative teaching methods, such as “food labs,” and use of a variety of carbohydrate resource tools and publications. In this article, glycemic effects of protein, fat, and fiber intake are discussed for persons with type 1 and type 2 diabetes. Decision trees are introduced for each level of carbohydrate counting and show the usual progression through each level. Carbohydrate counting as a meal planning approach offers variability of food choices with the potential for improving glycemic control. Research opportunities are available for those interested in comparing carbohydrate counting with other meal planning approaches for clients with diabetes and the effects on clinical outcomes. J Am Diet Assoc. 1998;98: 897-905.