AimUnderstanding and predicting ecosystem functioning such as biomass accumulation requires an accurate assessment of large-scale patterns of biomass distribution and partitioning in relation to climatic and soil environments.MethodsWe sampled above- and belowground biomass from 26 sites spanning 1500 km in Inner Mongolian grasslands, compared the difference in aboveground, belowground biomass and below-aboveground biomass ratio (AGB, BGB, and B/A, respectively) among meadow steppe, typical steppe, and desert steppe types. The relationships between AGB, BGB, B/A and climatic and soil environments were then examined.ResultsWe found that AGB and BGB differed significantly among three types of grasslands while B/A did not differ. Structural equation model analyses indicated that mean annual precipitation was the strongest positive driver for AGB and BGB. AGB was also positively associated with soil organic carbon, whereas B/A was positively associated with total soil nitrogen.ConclusionsThese results indicated that precipitation positively influence plant production in Inner Mongolian grasslands. Contrary to the prediction from the optimal partitioning hypothesis, biomass allocation to belowground increased with soil total nitrogen, suggesting that more productive sites may increase belowground allocation as an adaptive strategy to potentially high fire frequencies.