Wetting of dry soil triggers a pulse of microbial respiration that has been attributed to two broad mechanisms: (1) recycling of microbial cellular carbon ©, and (2) consumption of extracellular organic C made available to microbes by wetting. We evaluated these two mechanisms by measuring cumulative CO2 release, changes in the size and chemical composition of microbial biomass, and water-extractable organic carbon (WEOC) concentrations following artificial wetting of soil sampled from two depths at each of seven sites across California spanning a range of geologic parent materials. In samples collected from surface soil (0–10 cm depth), we found that cumulative CO2 release after wetting in the laboratory was most strongly correlated with microbial biomass. In these samples, the relative abundance of trehalose—a putative microbial osmolyte—decreased from 25% (SDþinspace=þinspace12) to 16% (SDþinspace=þinspace7) of the chloroform-labile fraction of the microbial biomass after wetting. This suggested a role for osmolyte consumption in generating the respiration pulse. In subsoil (40–50 cm depth, or sampled at contact with rock), however, the cumulative CO2 release after wetting was unrelated to microbial biomass and more strongly related to WEOC. The concentrations of selected microbial biomass constituents (e.g. trehalose and amino acids) in WEOC were negligible (<þinspace1%), suggesting that cell lysis was not important in generating WEOC in this study. The amount of WEOC relative to total organic C was greatest in subsoil, and negatively related to ammonium oxalate-extractable Fe (Pearson’s Rþinspace=þinspace0.42, pþinspace<þinspace0.01), suggesting a role for soil mineralogical properties in controlling WEOC release. Together, these findings suggest that microbial cellular C and extracellular C jointly contribute to the respiration pulse, and that their relative contribution depends on depth.