Leaf elemental stoichiometry has an important role in linking plant response to environmental changes and its afterlife effect. Plant leaf stoichiometry can change dramatically across environment gradients such as grazing, moisture and soil fertilities and consequently affect ecosystem function (e.g. nutrient cycle) in turn. Although concentrations of elements in leaves have already be known to have their effects on litter decomposability, the effects of multiple elements were less studied, especially other elements besides C, N, P. Since there are coordinated decomposition rates of different organs, it can be hypothesized that leaf stoichiometry could capture the variation of litter turnover time and therefore feedback to ecosystem function. To explore the effects of biogenic element concentration in leaves on litter turnover, biomass and litter aboveground were collected from 38 plots in a savanna grassland (E101 °35 '-102 °06 ', N25 °23 '-26 °06 ') which situated at the arid-hot valley of Jinsha river. Biomass and litter were classified into species before weighting. Litter turnover time was calculated as the fresh biomass : litter pool ratio. The most important biogenic element (i.e. N, P, K, Ca, Mg, S, Mn, Cu, Fe and Zn) concentrations in leaves were measured. The relationships between these element concentrations in leaves and litter turnover time were analyzed. The results showed that litter turnover may be affected by multiple elemental concentrations in leaves. K has a significantly positive relationship with litter turnover time while S, Mn and Mg were negatively related to litter turnover time which indicated K might inhibit litter decomposition while S, Mn and Mg might promote it. Contrary to previous studies, neither N nor P was significantly related to litter turnover time. K, S and Mn were related to litter turnover time significantly at species level respectively. K and S jointly explained 16.93% variation at species level. K, S, Mg and Mn were related to litter turnover time significantly at plot level respectively. Although N had no significant effect on litter turnover time alone, K-N and their interaction explained 37.42% variation at plot level. This combination explained 18.37% more variation than K-N additive effect. It was followed by K-Mn combination which explained 36.64% variation at plot level. Other elemental combinations could also significantly explained variation of litter turnover time to some extent. K was positively related to litter turnover time although it was positively related to Mn and Mg which were in turn negatively related to litter turnover time. The results here suggest that K might be the only element that inhibits litter decomposability in this dry and hot area. The mismatch of elemental effects on litter turnover time between at species level and at plot level might result from element transfer in litters or their complementary effects in decomposer's growth. The interactions of multiple elements indicated that there might be synergic and antagonistic effects between elements and suggested that litter turnover time would be affected by multiple decomposers which may be limited by different elements. The results here suggest that future studies should make attentions on other elements besides N and P to explore their roles in biogeochemistry processes.