Myrcia rhodeosepala was described in the Flora Brasiliensis by Berg (as Aulomyrcia lanuginosa) and transferred to Myrcia by Kiaerskou as the specific epithet ‘lanuginosa’ was exhausted in this genus. Myrcia rhodeosepala was accepted as a good species during the 20th and early 21st century, but in 2008, it appeared as a synonym of Myrcia tomentosa in The World Checklist of Myrtaceae. There is now convincing historical, geographical, morphological, anatomical and ecological evidence that M. rhodeosepala should be reinstated. First, the type, cited to be from Rio de Janeiro by Berg, was probably collected in Minas Gerais. Second, one of the publications cited as the source for the synonymy recognizes M. rhodeosepala as distinct from M. tomentosa. Morphologically, M. rhodeosepala can be distinguished by being a subshrub (v. shrubs or trees up to 10m in M. tomentosa), inflorescences sympodial throughout and forming showy synflorescences (v. in the axils of young leaves and distally monopodial in M. tomentosa), and leaf shape, size and venation. Anatomically the two species also differ. In M. rhodeosepala the adaxial leaf surface epidermal cells have straight anticlinal cell walls (v. sinuous in M. tomentosa). In M. rhodeosepala epicuticular wax occurs on both surfaces of the leaf forming a densely coalescent reticulum (v. only on the adaxial surface of the leaf in small, sparse granules in M. tomentosa). Finally, M. rhodeosepala produces significantly (T-test significance: p< 0.001) heavier flowers, fruits and seeds in smaller numbers per plant than M. tomentosa at peak flowering and fruiting.