Penelitian ini bertujuan untuk melihat pengaruh model pembelajaran learn on delivery (leader) terhadap hasil belajar siswa pada pembelajaran biologi kelas x di sma muhamadiyah boarding school tarakan. jenis penelitian yang digunakan adalah penelitian quasi eksperiment dengan desain non equivalent control group design. penelitian dilaksanakan pada bulan maret selama dua pertemuan dan sampel yang digunakan adalah x mia 1 (kelas eksperimen) dan x mia 2 (kelas kontrol). teknik pengumpulan data dengan observasi dan tes formatif berupa 30 soal pilihan ganda serta dokumentasi. uji normalitas menunjukkan nilai sig ≥ 0.05 yang berarti sampel berdistribusi normal, uji homogenitas menunjukkan nilai 0,185 dan 0,140 (sig ≥ 0,05) yang berarti data bersifat homogen. hasil penelitian menggunakan analisis of covariance (anakova) menunjukkan bahwa model pembelajaran learn on delivery (leader) berpengaruh terhadap hasil belajar sudah layak/benar dilihat dari nilai sig 0,000 (< 0,05) dengan nilai f sebesar 46.981 (corrected model) pada tabel (tests of between-subjects effects) nilai signifikan 0,000 (p 0,05). dengan demikian dalam model anakova ini peranan kovariat preetest tidak signifikan.

This research aimed to figure out the influence of learn on delivery (leader) learning model on biology learning outcomes of grade x students at sma muhammadiyah boarding school tarakan. the research employing quasi experimental method with a non equivalent control group design was conducted in march for two meetings with the samples of x mia 1 (experimental class) and x mia 2 (control class). the data were collected using observations, formative tests in the form of 30 multiple choice questions, and documentation. the normality test shows tha significance value was 0.05, meaning that the samples were normally distributed. the homogeneity tests shows the value of 0.185 and 0.140 (sig 0.05), meaning that the data is homogeneous. the research results based on the analysis of covariance (anacova) show that learning on delivery (leader) learning model which influenced learning outcomes was feasible seen from the sig value of 0.000 (< 0.05) with f value of 46.981 (corrected model) in the table (tests of between-subjects effects) with the significant value of 0.000 (p 0.05). thus, in this anacova model, the role of covariate in pretest was not significant.