The purpose of this article is to explore the process of children acquiring Japanese as first language from the perspective of Usage-based Model. The data is a child’s utterances which are collected from her age between 2;0-4;0. “ wo”, “ ga”, “ ni ” and, “ de ” -constructions are extracted from the data and qualitatively analyzed. According to the analysis, several features are found in her utterance, which are shown as below: 1 . The Japanese L1 child began to produce the item-based constructions around age of 2;0. 2 . With respect to the “wo” constructions, most of them appeared with omissions of the particle “wo”. These results are consistent to the hypothesis of Tomasello’s Usage-based model, which describes the process of L1 acquisition as “item-based construction”.