Penelitian ini bertujuan untuk mengembangkan perangkat pembelajaran model koopertaif tipe Two Stay Two Stray yang valid, praktis, dan efektif pada pokok  bahasan ekosistem kelas V; dan untuk mendeskripsikan hasil keterlaksaan RPP, hasil belajar siswa, keterampilan sosial, respon siswa, serta hambatan-hambatan yang dihadapi dalam pelaksanaan pembelajaran. Penelitian pengembangan perangkat pembelajaran model kooperatif Tipe Two Stay Two Stray yang mengacu pada model pengembangan four-D tetapi hanya sampai pada tahap pengembangan. Subjek penelitian ini adalah siswa kelas VA dan VB SDN Ngagelrejo I/396. Uji coba dilakukan dengan menggunakan One Group Pretest-Posttest Design. Pengumpulan data dilakukan melalui pengamatan, tes, dan angket respon siswa. Teknik analisis data menggunakan teknis analisis secara deskriptif kuantatif. Hasil penelitian ini menunjukkan bahwa (1) perangkat pembelajaran yang dikembangkan berkategori sangat valid dan reliabel dengan  rincian rata-rata validasi silabus 3,6 dengan reliabilitas 85,72 %, rata-rata validasi RPP 3,6 dengan reabilitas 90,42%, rata-rata validasi BAS 3,6 dengan reliabilitas 92,96%, rata-rata validasi LKS 3,8 dengan reliabilitas 85,72% dan rata-rata validasi THB 3,8 dengan reliabilitas 90,74%. (2) perangkat pembelajaran yang digunakan praktis dengan rata-rata presentase keterlaksanaan RPP VA 94,7% dan VB 84,3% serta keterbacaan BAS sebesar 96%. (3) perangkat pembelajaran yang dikembangkan efektif dalam meningkatkan hasil belajar. Hal ini dibuktikan dengan presentasi ketuntasan pada ranah pengetahuan 96%, ranah sikap 98% dan ranah keterampilan 100%. Respon siswa berkategori kuat dengan presentase 98%. Dari hasil penelitian ini dapat disimpulkan bahwa model kooperatif tipe Two Stay Two Stray dapat meningkatkan hasil belajar siswa kelas V SD.

This study aims to develop a valid, practical and effective Two Stay Two Stray type of cooperative learning model’s tools on the subject of fifth grade ecosystems and to describe the results of the implementation of lesson plans, student learning outcomes, social skills, student responses, and obstacles faced in the implementation of learning. Research on the development of cooperative learning model Type Two Stay Two Stray models which refers to the four-D development model but only reaches the development stage. The subjects of this study were VA and VB grade students at Ngagelrejo I / 396 SDN. The trial was conducted using One Group Pretest-Posttest Design. Data collection is done through observation, tests, and student questionnaire responses. The data analysis technique uses descriptive quantitative technical analysis. The results of this study indicate that (1) the learning tools developed are categorized as very valid and reliable with details of the average syllabus validation of 3.6 with a reliability of 85.72%, the average of RPP validation of 3.6 with a reliability of 90.42%, the average the BAS validation was 3.6 with a reliability of 92.96%, the average validation of LKS was 3.8 with a reliability of 85.72% and the average validation was THB 3.8 with a reliability of 90.74%. (2) learning tools that are used practically with an average percentage of implementation of RPP VA 94.7% and VB 84.3% and BAS readability of 96%. (3) learning tools developed are effective in improving learning outcomes. This is evidenced by the presentation of completeness in the 96% knowledge domain, 98% attitude domain and 100% skills domain. Student responses were categorized as strong with a percentage of 98%. From the results of this study it can be concluded that the Two Stay Two Stray cooperative model can improve student learning outcomes for fifth grade elementary schools.