PIP: Over the short term, irrigation projects often negatively influence the nutrition and health of target populations through factors which are either directly or indirectly related to the development process of the project. A cross-sectional survey of 187 households in, and 168 households from surrounding areas of the Mwea-Tebere rice irrigation scheme (MTIS) in Kenya was conducted to study the individual and social characteristics associated with childhood malnutrition. The cultivated area in the MTIS grew from 7000 acres in 1965 to 31,000 in 1987. However, during the same period, the population grew from 16,000 to 40,000, with most or all of the tenants' children continuing to live with their parents and overcrowding scheme villages. The proportion of under-fives who had weight/age, weight/height, and height/age indices under -2 standard deviations of the WHO-NCHS reference values was higher among MTIS residents than among nonresidents. Maternal education was the only variable significantly associated to all indicators of malnutrition used in the study. the area of residence, the number of resident dependents upon proceeds from the same farm, recent episode of diarrhea, child's age, mother's occupation, mother's age, type of caretaker of the child in the mother's absence, and mother's marital status were associated to one or two indicators of malnutrition. The level of maternal education is significantly higher at MTIS than off-scheme. However, the influence of MTIS' crowding and poor sanitation conditions is strong enough to make the nutritional status of preschool children worse within the scheme. Maternal education should be promoted, facilities provided to improve sanitary conditions, and programs launched to control population density in on-scheme villages.