The alpine meadow of the Tibetan Plateau is the fifth largest rangeland of China. The estimated population density of yaks is about 10/hm²; the recorded dung density is about 3600 pats/hm² on average and sometimes can reach 5900 pats per hm², occupying about 24% of the total grassland area in summer grazing pastures. The decomposition rate of yak dung is therefore of great importance to ecosystem primary productivity and nutrient cycling of the alpine meadow. Previous studies have demonstrated that the decomposition rate is largely dependent upon both abiotic factors such as temperature and rainfall and biotic factors including microbial biomass and the abundance of beetles, flies, and many other decomposers. Changes of the abiotic factors may affect the dung physical properties and dung the decomposer abundances, collectively influencing the rate of the dung loss rate. However, the effect of the abiotic factors on the dung decomposition has scarcely been examined in relation to the decomposer abundances. The amount of rainfall fluctuates greatly between years and the magnitude of the interannual variation may become greater in the future warmed world of the Tibetan Plateau. In this study, in order to determine the potential effect of rainfall regimes on the decomposition rate of yak dung and associated underlying mechanisms, we artificially manipulated rainfall amount to conduct an experiment involving three treatments including yak dung pats covered by plastic sheet (rainfall exclusion), sprayed with additional water (50% more than natural rainfall), and exposed to natural rainfall (as control) in an alpine meadow of northwestern Sichuan Province. For each treatment, there were 30 replicates that were sampled for 6 times (on the third, fifth, eighth, thirteenth, eighteenth, and twenty-fifth days after starting the experiment), each having five replicates. Water content, the number of coprophagous beetle adults and larvae that were hand-collected, and the dung dry weight were surveyed for 6 times, total nitrogen and phosphorus concentrations of the dung pats were measured for three times (On the starting day and the fifth and twenty-fifth days). Significant reductions in dung water content, weight loss, and the number of beetle adults were found in the treatment of rainfall exclusion, but rainfall addition did not change these variables possibly because of the high amount of rainfall during the experiment period, i.e. the rates of dung loss and nutrient release were indistinguishable between the control and the treatment with additional water. During the early stages of the experiment (particularly in the first five days), water content was positively correlated with the number of beetle decomposers, which was in turn positively correlated with the rate of dung loss. This indicates that the change of rainfall amount might have indirectly affected the dung decomposition rates via directly changing the decomposer abundance. However, the manipulation of rainfall amount did not lead to significant differences in total nitrogen and phosphorus concentrations of yak dung among the three treatments. This suggests that nutrient release from dung to soil was proportional to the dung loss during the experiment. In conclusion, reduction of rainfall amount may slow down dung decomposition and nutrient cycling in the study alpine meadow ecosystems and hence perhaps reduce the ecosystem primary productivity over the long run.