The problem of high rates of child malnutrition remains in Pagar Alam City. The lack of understanding of nutritional interventions and the limited ability of Posyandu cadres to conduct accurate nutritional assessments are the main factors. This situation makes it difficult for the community to monitor the nutritional status of children and provide appropriate nutritional intake. This research aims to create an intelligent system for detecting children's nutritional status through the exploration of C.45 algorithm and Forward Feature Selection in Pagar Alam City. This system is designed to detect children's nutritional status and provide recommendations for appropriate nutritional intake based on detection results with variables of children's weight and height. The data used amounted to 7519 data obtained from the Pagar Alam City Health Office. The model we use to build this system is waterfall with stages of planning, analysis, design, development, testing and implementation. Then the method we apply to this system is CRISP-DM and the C.45 algorithm and Forward Feature Selection technique. Our results are in the form of an intelligent system for detecting children's nutritional status, with the results of system testing using test data and training data showing 100% accuracy. In addition, black box testing also proves that the system works well as expected.