This study compared the efficacy and tolerability of three enteral formulas in critically ill patients with COVID-19 who were ventilated and in the prone position: (a) immunomodulatory (IMM), (b) ω3 and (c) maltodextrins (MD). Primary outcome was the percentage of patients who received both 80 % of their protein and calorie targets at 3 d after enrolment. Secondary, mechanical ventilation-free time, ICU mortality and markers of nutritional status. Tolerance of enteral nutrition was evaluated by diarrhoea and gastroparesis rate. A total of 231 patients were included, primary outcome achieved was in ω3 group (76·5 % v. 59·7 and 35·2 %, P < 0·001) v. IMM and MD groups. Mechanical ventilation-free time was longer in ω3 and MD groups: 23·11 (sd 34·2) h and 22·59 (sd 42·2) h v. 7·9 (sd 22·6) h (P < 0·01) in IMM group. Prealbumin final was 0·203 ± 0·108 g/L and 0·203 ± 0·095 g/L in IMM and ω3 groups v 0·164 ± 0·070 g/L (p < 0·01) MD group. Transferrin were 1·515 ± 0·536 g/L and 1·521 ± 0·500 g/L in IMM and ω3 groups v 1·337 ± 0·483 g/L (p < 0·05) MD group. Increase of lymphocytes was greater in ω3 group: 1056·7 (sd 660·8) cells/mm3 v. 853·3 (sd 435·9) cells/mm3 and 942·7 (sd 675·4) cells/mm3 (P < 0·001) in IMM and MD groups. Diarrhoea and gastroparesis occurred in 5·1 and 3·4 %, respectively. The findings of this study indicate that enteral nutrition is a safe and well-tolerated intervention. The ω3 formula compared with IMM and MD did improve protein and calorie targets.