Blenderized diets are not recommended because of difficulties in providing a constant content of nutrients, phoblems with viscosity and osmolarity, time needed for preparation, and increased risk of contamination. To determine the possibility of substituting commercial formulas for blenderized diets, 27 unconscious patients were randomized into two different groups : blenderized diet group(n=11) and commercial formula group(n=16). Blenderized diets were composed of food sources such as rice, milk, egg, juices and others, and its energy percentage of protein, fat, and carbohydrate was 15%, 30%, and 55%, respetively. Commercial enteral formulas provided 1 kcal/ml and the energy percentage of protein, fat and carbohydrate was 25%, 20% and 55% in greenbia(G) and 20%, 20%, and 60% in Greenbia-DM(GD), repectively. In commercial formula group, five patients with diarrhea, constipation, and high blood glucose level received GD, and the other eleven patients received G. All patients received their lipuid diet for 4 weeks through 16 French rubber nasogastric tube by bolus feeding 6 times daily. The patients in commercial formula group had a similar mean daily calorie intake to the patient in blenderized diet group, 32-34kcal/kg/d and 30-35kcal/kg/d, respectively. Patients given commercial formula gained an averge of 1.7kg in weight but those given blenderized diet lost an average of 2.6kg. There was a trend toward an increase in percent ideal body weight in commercial formula group(94$pm$5 vs 99$pm$6%). However, blenderized diet group showed a trend toward a decrease in percent ideal body weight(106$pm$5 vs $101pm$6%). The initial levels of total lymphocyte counts, serum albumin and tranferrin were below the normal rante, The mean daily protein intake was significanlty higher for commercial formula group(2.0-2.1g/kg/d) than for blenderized diet group(1.1-1.3g/kg/d). Patient fed cormmercial formula for 4 weeks showed a trend toward an increase in serum albumin(8.1%) and a significant increase in serum transferrin(32.1%) without increasing the levels of blood glucose. GOT, GPT, blood urea nitrogen and serum creatinine. In the blenderized diet group, however, no significant improvement were obseved in the concentration of serum albumin and transferin, compared to initial value. There was a trend toward a decrese in the level of hemoglobin and hematocrit in blenderized deit group but no significnat change in commercial formulas goups. All patients tolerated both diets well and no significant complications were encountered. The results indicate that the commercial enteral formulas tested in this study can be an effective substitution for blenderized diet in unconscious tube-fed patients.