Physiologic responses of 30 enterally-fed long-term care residents with type 2 diabetes receiving total nutrition support via either a disease-specific (reduced-carbohydrate, modified-fat) formula or a standard high-carbohydrate formula for 3 mo were compared. Objectives of the study included evaluating metabolic response (glycemic control and lipids) and clinical outcomes. Thirty-four subjects requiring total enteral nutrition support by tube were enrolled in this prospectively randomized, double-blind, controlled, parallel group 3-mo pilot trial. Thirty were evaluable in that they completed 4 wk. Twenty-seven completed all 12 wk. The groups were well-matched for physiologic and demographic parameters at baseline. Fasting serum glucose and capillary (fingerstick) glucose values demonstrated better control in the disease-specific formula-fed group. Serum lipid profiles of this group were similar to or better than those of the standard formula-fed group. The amount of insulin administered to insulin-using subjects in the disease-specific formula-fed group was consistently less than before initiation of the formula, whereas the amount administered was consistently higher in the group fed the standard formula. Overall, subjects randomized to the disease-specific formula experienced better numerical biochemical control and better clinical outcomes when expressed on a numerical and percentage basis. These included surrogate markers of diabetes control such as serum glucose and glycohemoglobin, as well as clinical outcomes such as incidence of infections and pressure ulcers. These findings confirm that the disease-specific formula provides better glycemic control, poses no risk to lipoprotein metabolism, and provides for better clinical outcomes.