In arid and semi-arid ecosystems, recreational activities are largely confined to riparian forests due to their relatively favorable microclimate and access to water. This study aimed to assess the impact of human recreational activity on carbon stocks in riparian forests (Khuzestan Province, Iran). Specifically, it evaluated the effects of such activities on Populus euphratica stands and examined changes in soil nutrients resulting from recreational use. To achieve this, two types of pure Populus euphratica stands were selected: one group exposed to recreational human activity, and another located in undisturbed areas, distant from recreational use. Within each group, four stands were chosen, and a sampling plot measuring 30 x 30 m was established in each stand. Soil samples were collected to a depth of 15 cm, and carbon reserves were calculated based on this soil layer. Particulate carbon stock (PCS; carbon in soil fractions >0.053 mm) and mineral-associated carbon stock (MACS; carbon in soil fractions <0.053 mm) were determined. Soil carbon stock was significantly higher in non-recreational areas (55.86 t ha(-1)) compared to recreational areas (14.83 t ha(-1)). Non-recreational areas exhibited higher particulate carbon stock (PCS; 13.09 t ha(-1)) and mineral-associated carbon stock (MACS; 42.77 t ha(-1)), whereas recreational areas showed lower PCS (8.46 t ha(-1)) and MACS (6.36 t ha(-1)). Biomass carbon stocks were 49.9 t ha(-1) in recreational areas and 87.7 t ha(-1) in non-recreational areas. Soil nitrogen content was 0.04% in recreational areas and 0.17% in non-recreational areas. Intense recreational activities led to approximately 73 and 57% reductions in soil and biomass carbon stocks, respectively, when compared to non-recreational areas. These results indicate that the riparian site is being used improperly and beyond its ecological carrying capacity. Overall, intensive recreational use of the riparian forest contributed to pronounced soil degradation and a marked reduction in carbon sequestration potential.