Background: In vitamin A-deficient populations, children hospitalized with infections and/or malnutrition are at particular risk of developing severe vitamin A (VA) deficiency. High-dose VA supplements are recommended as part of the treatment but results on its effect on recovery from morbidity and on prevention from nosocomial morbidity are conflicting. Objective: We aimed to assess the effect of a single high dose and daily low dose of VA on hospitalized malnourished children's morbidity. Design: We carried out a double-blind, randomized trial in 604 and 610 Senegalese hospitalized children. The first mentioned batch received a high-dose VA supplement (200 000 IU) on admission, the second a daily low-dose VA supplement (5000 IU per day) during hospitalization. Children were followed up until discharged. Data on all-cause morbidity were collected daily. Results: Survival analysis showed that the incidence of respiratory disease was significantly lower in the low-dose group than in the high-dose group, hazard ratios (HR): 0.26, 95% CI: 0.07-0.92. The duration of respiratory infection was also significantly lower in the low-dose group than in the high-dose group (HR of cure: 1.41, 95% CI: 1.05-1.89). Duration and incidence of diarrhoea were not significantly different between treatment groups. In children with oedema on admission, mortality was significantly lower in the low-dose group (Adjusted odds ratio: 0.21; 95% CI: 0.05-0.99). Conclusions: Daily low dose of VA compared with single high dose significantly reduced duration and incidence of respiratory infection but not of diarrhoea in hospitalized children.