Mountains hold large but poorly recognized carbon stocks, yet the carbon-market rules that finance afforestation, reforestation and revegetation (ARR) were developed for lowland landscapes. Following PRISMA 2020, this systematic review examined whether eligibility, monitoring, permanence and governance requirements work in mountain terrain. Four bibliographic databases, institutional repositories and carbon-standard registries were searched for 2000–2026. Of 5040 records, 185 sources were included, covering five mountain regions, with the Indian Himalayan Region as the principal case. Carbon stocks varied strongly with elevation, aspect and previous land use, and much was stored in soils and non-forest vegetation. Assisted natural regeneration and mixed native planting outperformed exotic monocultures on carbon durability, biodiversity, water yield and livelihoods. In the Andean páramo, pine plantations reduced surface soil organic carbon from 5.0 to 3.5 kg m⁻² and halved water yield. Measurement accuracy declined on steep slopes, and default buffer pools appeared to underestimate reversal risk. The evidence suggests that the main constraint is regulatory design rather than biophysical potential, because current rules favor large single-owner plantations. We propose four regulatory reforms and a seven-step pathway for mountain project development.