Mangrove vegetation can be easily recognized from remote-sensing satellite images compared to other terrestrial vegetation. The vegetation index is applied to the satellite images to highlight the aspect of vegetation density. This study aims to determine the correlation between the value of the vegetation index and mangrove canopy cover data to achieve a proper vegetation index to estimate the density of the mangrove canopy. The data needed are satellite imagery from Landsat 8 and mangrove canopy cover in sampling locations along the coast of Dodinga Bay, West Halmahera. Image data analysis includes radiometric correction, image sharpening, masking, classification, and accuracy tests. The vegetation index algorithms used were NDVI, GNDVI, and IM, and regression analysis was carried out for correlation tests. The analysis results obtained four different land cover classes with an overall accuracy of 97.70% and a kappa coefficient of 0.9688. The IM vegetation index showed an excellent correlation with mangrove canopy cover compared to GNDVI and NDVI. The determination coefficient (R²) of the IM is 0.6765; GNDVI (0.4897) and NDVI (0.4825). The IM classification produces four levels of mangrove canopy density, i.e., sparse (7.40 ha), moderate (628.33 ha), dense (921.22 ha), and very dense (16.45 ha). Keywords: mangrove, Landsat 8 images, vegetation index, Dodinga Bay Abstrak Objek vegetasi mangrove paling mudah diidentifikasi dengan menggunakan citra satelit penginderaan jauh dibandingkan objek vegetasi darat lainnya. Indeks vegetasi diterapkan terhadap citra untuk menonjolkan aspek kerapatan vegetasi. Penelitian ini bertujuan mengetahui korelasi antara nilai indeks vegetasi dengan data tutupan kanopi mangrove, sehingga didapatkan indeks vegetasi yang sesuai untuk menduga kerapatan kanopi mangrove. Data yang diperlukan yaitu citra Landsat 8 dan tutupan kanopi mangrove di lapangan. Analisis data citra terdiri dari koreksi radiometrik, penajaman citra, masking, klasifikasi dan uji akurasi. Algoritma indeks vegetasi yang digunakan yaitu NDVI, GNDVI dan IM, serta dilakukan analisis regresi untuk uji korelasi. Hasil analisis mendapatkan empat kelas tutupan lahan yang berbeda dengan overall akurasi sebesar 97,70 % dan kappa coefisien sebesar 0,9688. Indeks vegetasi IM menunjukkan korelasi sangat baik dengan tutupan kanopi mangrove dibandingkan GNDVI dan NDVI. Koefisien determinasi (R²) IM adalah 0,6765; GNDVI (0,4897) dan NDVI (0,4825). Klasifikasi IM menghasilkan empat tingkat kerapatan kanopi mangrove yaitu mangrove jarang (7,40 ha), mangrove sedang (628,33 ha), mangrove lebat (921,22 ha), dan mangrove sangat lebat (16,45 ha). Kata kunci: mangrove, citra Landsat 8, indeks vegetasi, Teluk Dodinga