Malnutrition refers to a condition that occurs due to insufficient intake of micronutrients and macronutrients necessary to support optimal bodily functions in toddlers. The World Health Organization (WHO) reported in 2022, 149 million children under five experienced stunting, 45 million suffered from wasting, and 37 million were classified as overweight or obese. Despite the availability of anthropometric data, predictive analysis remains underutilized, hindering early interventions. Machine learning was applied to classify toddler nutritional status using 3,130 records from Kalijambe Health Center. The performance of Decision Tree and Support Vector Machine (SVM) models was compared based on anthropometric indices: Weight/Age, Height/Age, and Weight/Height. Preprocessing involved handling missing data, standardizing features, and addressing class imbalance. The results showed that Support Vector Machine (SVM) demonstrated superior performance, achieving the highest accuracy of 93.2% in Weight-for-Age (W/A), compared to 79.5% for Decision Tree. SVM was more effective in handling class imbalance and produced a more stable model. The approach provides a more reliable framework for predicting malnutrition risks, potentially guiding targeted toddler health interventions in Indonesia.