Children under five are more at risk of experiencing nutritional disorders because during this period children will experience very rapid growth and development. Nutritional disorders can result in children's performance in activities not being optimal, hampering the growth and development process, and can even cause disease. Malnutrition sufferers in Indonesia are increasing every year, and the percentage of malnutrition in Indonesia is around 3.4%. An expert system is a computer-based system built based on facts, knowledge and reasoning that can help solve a problem. However, to measure the uncertainty and level of expert confidence in identifying and detecting malnutrition, an expert system can use the Dempster-Shafer method. The results of testing the system model using the Dempster-Shafer method on toddler nutritional diseases obtained quite good results with system accuracy of 84% with a precision value of 81.83% and also a recall value of 80%.