In the Kivu region located in east of the Democratic Republic of the Congo, malnutrition and malaria is a major cause of morbidity and mortality. The relationship between malaria and malnutrition is unclear and has never been studied in the Kivu region. This report presents an analysis of data from 5695 children aged 0 to 5 years, admitted to the paediatric ward of Lwiro hospital between November 1992 and February 2004. The weight/age (W/A) index and weight/height (W/H) index expressed with standard deviation in relation to the reference median were calculated (Z score). The association between protein-energetic malnutrition and malaria infection and nutritional indicators was measured based on prevalence ratios determined by univariate analysis and adjusted Odds Ratio (OR) derived using a multivariate model. The prevalence of malaria at the time of admission was 35.8 % (n=5695). The W/A and W/H indexes and serum albumin level were correlated with malaria-related morbidity. Logistic regression showed that high malaria OR was associated with both anthropometric nutritional indicators [WHZ > -2: OR (CI 95 %) 1.7 (1.4-2.2)] [WAZ > -2: OR (CI 95 %) 1.3 (1.1-1.6)] and biological nutritional indicators [serum albumin > or = 23 g/L: OR (CI 95 %) 1.6 (1.2-2.1)]. Our findings indicate that malnourished children at admission have a lower risk of malaria infection.