As an important surface cover in drylands, biocrusts play a critical role in regulating regional carbon (C) emissions through respiration. However, the partitioning of biocrust soil respiration into its components (such as moss (R-moss) and microbial respiration (R-microbial)) and their relative contributions to C efflux remain unclear. In this study, we conducted continuous measurements of respiration rates on moss crusts with different coverage (0, 5%, 25%, 50%, 75%, 100%) over two growing seasons (2023-2024) on the Loess Plateau in northern China. Our results showed that the biomass regression method effectively partitioned R-total into R-moss and R-microbial (R-2 = 0.28-0.99), while a coupled soil water content-temperature model explained 65% and 41% of their variation, respectively. Soil water content showed an indirect positive association with R-microbial via soil organic C content (P < 0.001), whereas soil temperature showed an indirect negative association with R-moss via soil water content and moss biomass (P < 0.001). Model extrapolation estimated that an annual C efflux from moss crust was 564 g C m(-2) yr(-1), with R-moss and R-microbial contributing 51% and 49%, respectively. Overall, our study reveals that biocrust respiration components (R-moss and R-microbial) contribute approximately equally to total C emissions and highlights their partitioning as critical for improving global dryland C cycle projections.