This research aims to produce valuable and practical edutainment learning model guidelines. This research method refers to the theory of Borg and Gall. The research steps are as follows: information gathering, research planning, developing initial products, field testing, revisions, product testing, and revision of field product test results. Data collection techniques used are interviews, observations, and questionnaires. The analysis used in this research is quantitative and qualitative. The data sources in this study were the teachers and students of Raudhatul Athfal group B in Kudus. The results of this study are described in a needs analysis, model design, feasibility test, and effectiveness test. Needs analysis shows a need for an appropriate learning model to improve early childhood language skills. This research uses research and development (R&D) methods. The data analysis techniques used are model feasibility analysis and effectiveness. The edutainment learning model is a learning model that is carried out in the process to create a pleasant learning quality for students. So the nature-based edutainment learning model is used. The model design uses its products' ten research steps: learning models and nature-based edutainment learning guides. Feasibility test with N Gain test, which shows an average experimental class of 0.71 and 0.72 with a high Gain Classification with an "effective" level of effectiveness (g>0.70). In addition, the feasibility test is also carried out with the Expert Judgment Test consisting of language and material experts. The language and content experts' assessments obtained an average validity score of 80% and 88% ("very good" category). Test the effectiveness with homogeneity, normality, and t-tests. In the homogeneity test, it is known that the significance value of children's language skills is 0.151 > the value of = 0.05, then H0 is accepted. The normality test obtained a significance value for children's language skills, which is more than 0.05, so it is normally distributed. The t-test, the language skills of children in group B using the t-test obtained a significance value (2 tailed) of 0.000 <0.005.