OBJECTIVES: Worldwide childhood undernutrition is a major risk factor for death and disability. La Guajira (Colombia) is a mostly rural state, with the highest proportion of indigenous population. The aim of this study was to estimate the factors related to undernutrition in under-five population (U5P) in one of the regions with the highest burden of disease due to undernutrition in Colombia. METHODS: We designed a cross-sectional study with data from the National Survey of the Nutritional Situation in Colombia (ENSIN, 2010), to identify the undernutrition associated factors in U5P. The outcome variable was undernutrition, defined as weight-for-age lesser or equal than 2 standard deviation (SD). Odds Ratios (OR) adjusted by taking into account the muestral design were calculated to assess the risk of each group compared to a reference group. Statistical analyzes were performed in STATA 12. RESULTS: A total of 622 children under-five were analyzed. The prevalence of undernutrition was 11.2%, 58.3% were male, 79.1% lived in rural areas, 36.2% do not have health insurance, and 73.1% were indigenous. According to the birth order, 13% were the first, 26% were the 2-3, 25% the 4-5 and 36% the 6 or more. In the logistic model, children living in houses with sewage had a lower risk of acute undernutrition (OR, 0.16; 95% CI, 0.05-0.57), while being the sixth child or later increases the risk of undernutrition (OR, 4.07; 95% CI, 1.50-10.99). CONCLUSIONS: Community-based interventions are needed to reduce the high prevalence of undernutrition in La Guajira, Colombia. Priority should be given to nutritional and other socioeconomic interventions for large families. Being able to identify high risk households will increase the efficiency of the policies to prevent and control undernutrition in La Guajira.