Childhood is a golden age for children, which is important phase for their growth and development. Children need to obtain get a good education according to their development phase to improve their abilities including language. Early reading skill is considered as one type of language skills which is essential at the childhood period. Referencing to teacher's information, one of the problems related to students' reading ability is the absence of determination of the classification of early childhood reading skills. Currently, teachers are challenged in providing the achievement of students' reading skills based on criteria of early reading ability. Consequently, they are not able to improve the learning quality to each student nor prepare material to discuss with students' parents. There is a suggestion given to achieve the goals and resolve it. It is the classification of early childhood reading skills that has been designed where the data processing is carried out using a computer. This study employs a decision tree and Algorithm C4.5 for the method. Algorithm C4.5 is included in the classification based on Decision Tree. Algorithm C4.5 is a method that can handle features with numeric types. The purpose of this study is to perform classification using Algorithm C4.5 as a method that can find out the level of reading skills in early childhood, based on the criteria, so that classification can be done easily and get the right improvement suggestion. The result in this study is that the Algorithm C4.5 method has a good result in solving and calculating for the classification so that the accuracy value of the Algorithm C4.5 calculation that has been carried out is 77,14%.