Carbon use efficiency (CUE) describes the relative partitioning of carbon (C) between anabolic and catabolic processes within the soil microbial community. Further, it represents a major factor regulating the amount of C cascading through the trophic levels of the soil food web. How CUE relates to C supply, however, remains poorly understood. The primary aim of this study was to determine how CUE varies across a range of spatial scales as a function of C substrate supply. Our secondary aim was to understand how variations in substrate CUE influences the interpretation of community level physiological profiles (CLPP). Using 16 different 14C-labelled substrates (including amino acids, sugars, organic acids and amino sugars) and soils collected at the field, regional and continental scale, we measured the rate of substrate uptake and mineralization from which we calculated CUE. Across all soils (n = 114) and substrates (n = 16), the average CUE for the microbial community was 0.568 ± 0.004 (range 0.492–0.794). While the partitioning of substrate-C within the biomass (immobilization/mineralization) over 72 h was highly conserved for some substrates (e.g. glucose), others showed a wide variability in CUE across the samples (e.g. valine). In the context of the CLPP methodology, we showed that individual sites could be statistically separated from each other, irrespective of whether the statistical analysis was based on microbial substrate uptake rate or mineralization rate. However, our results do suggest that caution is needed when ascribing observed CLPP differences to the importance of individual C pathways operating in soil due to the wide variation of CUE between substrates. In conclusion, we present new mechanistic evidence to support the paradigm that variation in ecosystem CUE may in part reflect differences in the types of C supplied to the microbial biomass.