Background. Malnutrition is a global health problem. Madagascar has had an increase in the percentage of malnourished children since the beginning of 1980. Forty-nine percent of children less than 5 years of age are stunted. Research have shown that stunting is a result ofchronic malnutrition due to inadequate diet and recurrent illness.The political situation in Madagascar has changed towards a more democratic and transparent system. This may lead to more external donors and humanitarian aid. Both politicians and organisations will need updated research in order to design interventions in projects aimed at improving children’s health. The findings of this study will provide useful insight and local knowledge in infant feeding practices and management of diarrhoeal episodes.Objectives. The broad objective was to identify factors associated with malnutrition in young children, with an emphasis on knowledge and child rearing practices related to feeding andmanagement of acute, watery diarrhoea.Design and subjects. Cross-sectional survey was conducted on 143 households with children between 6 and 36 months. This is also the total number of subjects. In four villages all children in the age group were included (n=38). Another 105 children were included from two villages, interviewed consecutively. All the caretakers were interviewed based on a pre-tested, structured questionnaire with closed and open ended questions. Body weight, height and MUAC were measured for all children by trained personnel.Results. At the time of the survey 61.5% of the children were below -2SD WA, 50% were stunted and 9.9% were wasted. Lastly 12.6% were malnourished following MUAC. Mean duration of exclusively breastfeeding was 5.0 months (±2.1), mean duration of breastfeedingwas 20.2 months (±2.1). Fifty-six percent of the caretakers had an adequate knowledge in exclusively breastfeeding. Ninety-five percent were agreed that a child should start with weaning food at 6 months of age. However, 74% claimed that certain food should not begiven during introduction of complementary food. Mean age of introduction of weaning food was 5.6 months and the mean number of meals per day was 3.1. Thirty percent of the children had experienced diarrhoea within two weeks prior to the interview. Sixty-nine percent of the caretakers knew the symptoms of acute diarrhoea and99% knew more than four causes of diarrhoea. In 29 of the 43 cases with diarrhoea improper measures were taken. Eighty-four percent reported to have used ORS. Caretakers tended to mix symptoms and consequences of dehydration or include wrong symptoms 50% and45%,respectively. Fifty-eight percent agreed that a child need extra food after an episode with diarrhoea. Sixty-four percent of the caretakers reported that they boiled the child’s drinking water and 39% claimed that they always washed the child hands before meals. No significant association were observed between children’s nutritional status and the reported knowledge and practices. However, association were found between nutritional status and socio-economic and demographic factors as well as between the knowledge and rearing practices and socio-economic and demographic factors.Conclusion. Childhood malnutrition is a severe problem in this are. The infant feeding practices and management of diarrhoea were not optimal according to WHO’s recommendations. However, our results show a positive shift towards better feeding practices and rearing practices, especially among younger mothers. Even though this study reveal few specific risk factors for developing malnutrition, improvements in socio-economic strata is decisive as this is associated with some of the potential indicators of malnutrition. Further studies should focus on barriers in preventing caretakers from practicing their knowledge. Both governmental and non-governmental organisations should continue to promote andencourage caretakers in recommended practices.