Partitioning sources of ecosystem and soil respiration (R-eco and R-s) is important for understanding how climate change affects carbon cycling. Plant and microbial biomass analyses and daytime measurements of R-eco and R-s were performed for 25 plots in an alpine meadow at elevation 4313 m on the Tibetan Plateau. Plant and microbial biomass were determined by harvesting method and the chloroform fumigation-extraction method, respectively. Respiration fluxes were measured by an automated CO2 flux system (LI-8100, LI-COR Biosciences, Lincoln, NE, USA). Soil respiration can be estimated by a linear or exponential relationship between R-eco and aboveground plant biomass (AGB). Microbial respiration (R-m) can be estimated by a linear or exponential relationship between R-s and belowground plant biomass (BGB) or by a multiple relationship between R-eco and AGB and BGB. Soil respiration (or R-m) is respiration flux when AGB (or BGB) is extrapolated to zero for the linear and exponential regression methods. Similarly, R-m is respiration flux when both AGB and BGB are zero for the multiple regression method. Our findings suggest that the exponential regression method to partition sources of R-eco and R-s may be more appropriate compared to other methods for this alpine meadow of Tibet.