Abstrak: Penelitian ini dilatarbelakangi temuan pada observasi yang telah dilakukan peneliti bahwa di kelas I Madrasah Ibtidaiyah Islamiyah Mojokampung Bojonegoro terdapat beberapa siswa belum mampu membaca dengan lancar, berdasarkan permasalahan tersebut peneliti tertarik mengembangan media Puzzle Huruf Abjad dalam meningkatkan kemampuan membaca. Penelitian ini bertujuan untuk mendeskripsikan pengembangan serta mengetahui kelayakan media puzzle huruf abjad dalam meningkatkan kemampuan membaca kelas I. Penelitian ini menggunakan model penelitian dan pengembangan model ADDIE (Analysis, Design, Development, Implementation, Evaluation). Berdasarkan uji validitas yang telah dilakukan oleh peneliti Uji validasi dari ahli materi memperoleh persentase 92%, validasi ahli media memperoleh persentase 84% yang berarti media yang dikembangkan sesuai dengan materi dan layak untuk digunakan. Selanjutnya uji coba media Puzzle Huruf Abjad di kelas I mendapatkan nilai rata-rata pre-test 54 dengan ketuntasan klasikal 20%, sedangkan post-test memperoleh rata-rata 75,5 dengan ketuntasan klasikal 80% hal ini menunjukkan bahwa pengembangan media puzzle huruf abjad layak dan bisa digunakan dalam pembelajaran membaca permulaan.Abstract: This research was motivated by the findings from observations done by the researchers that there were several students who were not able to read fluently at Madrasah Ibtidaiyah Islamiyah (Islamic Elementary School) Mojokampung Bojonegoro, Class One. Based on the problem, the researchers sought to develop Alphabets Puzzle media to improve students’ reading skills. This study aims to describe the development and determine the feasibility of alphabetic puzzle media in enhancing the first graders’ reading skills. This research employed the ADDIE research and development model (Analysis, Design, Development, Implementation, Evaluation). Based on the validity test that has been carried out, the material expert validation test results obtained a percentage of 92%, followed by the media expert validation test results as much as 84%. It means that the media developed is in accordance with the material and is suitable for use. Furthermore, the implementation results of the learning media at Class I showcased an average pretest score of 54 with 20% classical completeness, while the post-test obtained an average of 75.5 with 80% classical completeness. This indicates that the development of alphabets puzzle media is considered feasible and can be used in learning to read.