The purpose of this study was to assess the effect of nutritional supplementation on dietary intake and on pressure ulcer development in critically ill older patients. The multi-center trial involved 19 wards stratified according to specialty and recruitment for critically ill older patients; 9 wards were randomly selected for nutritional intervention (nutritional intervention group), consisting of the daily distribution of two oral supplements, with each supplement containg 200 kcal, for 15 d. Pressure ulcer incidence was prospectively recorded for grades I (erythema), II (superficial broken skin), and III (subcutaneous lesion) for 15 d. Nutritional intake was monitored by using estimates in units of quarters validated by comparison with weight measurement. There were 672 subjects older than 65 y, and 295 were in the nutritional intervention group versus 377 in the control group. The patients were similar for age, sex ratio, and C-reactive protein. In comparison with the control group, the nutritional intervention group included more patients with stroke, heart failure, and dyspnea and fewer with antecedent falls, delirium, lower limb fractures, and digestive disease. The nutritional intervention group had a lower risk of pressure ulcers according to the Norton score but was less dependent (Kuntzman score) and had a lower serum albumin level. During the trial, energy and protein intakes were higher in the nutritional intervention group (day 2: 1081 +/- 595 kcal versus 957 +/- 530 kcal, P = 0.006; 45.9 +/- 27.8 g protein versus 38.3 +/- 23.8 g protein in the control group, P < 0.001). At 15 d, the cumulative incidence of pressure ulcers was 40.6% in the nutritional intervention group versus 47.2% in the control group. The proportion of grade I cases relative to the total number of cases was 90%. Multivariate analysis, taking into account all diagnoses, potential risk factors, and the intra-ward correlation, indicated that the independent risk factors of developing a pressure ulcer during this period were: serum albumin level at baseline, for 1 g/L decrease: 1.05 (95% confidence interval: 1.02 to 1.07, P < 0.001); Kuntzmann score at baseline, for 1-point increase: 1.22 (0.32 to 4.58, P = 0.003); lower limb fracture: 2.68 (1.75 to 4.11, P < 0.001); Norton score < 10 versus > 14: 1.28 (1.01 to 1.62, P = 0.04); and belonging to the control group: 1.57 (1.03 to 2.38, P = 0.04). In conclusion, it was possible to increase the dietary intake of critically ill elderly subjects by systematic use of oral supplements. This intervention was associated with a decreased risk of pressure ulcer incidence.