PURPOSE: To develop activity classifiers based on accelerometer data to recognize toddler’s eight distinct activities: walking/running, climbing up/down, standing, crawling, sitting, lying down, being carried, and riding a stroller/wagon. METHODS: Twenty-four toddlers aged 13 to 35 months (50% girls) performed various prescribed activities during free play in a commercial indoor playroom, while wearing Actigraph wGT3X-BT accelerometers on the hip and wrist. Their activities were video recorded. The video data were annotated and synchronized with accelerometer data. Five machine learning classifiers, including random forest, support vector machine, decision tree, K-nearest neighbors, and logistic regression, were trained and tested. Classifier performance was evaluated using subject-wise cross-validation. RESULTS: Activity classifiers were developed based on 1,011 two-second window accelerometer signal clips from the 24 participants. Of the five classifiers tested, the random forest classifier presented the highest overall accuracy (69% for hip and 55% for wrist). Overall, hip data showed higher accuracy than wrist data. Based on the hip random forest classifier, 91% of “walking/running” activities and 84% of “sitting” activities were correctly identified. However, 35% of “being carried” activities and 30% of “standing still” activities were misclassified as “walking/running”. Only 8% of “stroller/wagon ride” activities were misclassified as “walking/running”. CONCLUSIONS: This pilot study demonstrates that the machine learning approach can be used to detect toddler’s “walking/running” activities at a high level of sensitivity. However, the algorithm developed in this pilot study often misclassified “standing still” or “being carried” as “walking/running”. “Stroller/wagon ride” was less frequently misclassified as “walking/running”. Overall, hip data demonstrated higher accuracy than wrist data in detecting key activities for toddlers. Future research should follow to refine the algorithms and test external validity.