Small soil cubes of dimensions 12 by 12 by 12 mm were collected from the surface of a red earth. Treatments were addition of clover substrate or urea, drying and rewetting, or no amendment, after which soils were incubated at either –10 or –30 kPa. Each soil cube was analysed for NO-3 -N, NH+4 -N, total soil N (%N), volumetric water content (θv), microporosity (volume of pores <0·6 µm), and Mn 2+ concentration. Multiple regression analysis was used to determine if microporosity and Mn 2+ contributed uniquely to linear models in which %N and qv were also used to predict N mineralisation and nitrification. In soils incubated at –10 kPa, both microporosity and Mn 2+ had a strong positive influence on N mineralisation and nitrification, whereas in soils incubated at –30 kPa no such influence could be observed. These and other observations suggest that when soils with high microporosity were incubated at –10 kPa, O2 supply to the microbial biomass was limited and the reduction of Mn oxides to divalent Mn was enhanced. Increased substitution of Mn oxides for O2 as terminal electron acceptors in the microbially mediated oxidation of carbon substrates considerably increases H+ consumption. We propose that in the wetter soil (–10 kPa), this process relieves pH stress experienced by N mineralising and nitrifying organisms, thereby increasing their activity, but that in the drier soil (–30 kPa), O2 diffusion is less restricted and this mechanism does not operate appreciably. The influence of microporosity on clover-amended soils was to decrease levels of mineral N and this was attributed to greater denitrification in soils with high microporosity. Neither microporosity nor Mn2+ was an important variable in the prediction of mineral N in the urea-treated soils. This work highlights the interaction of physical, chemical, and biological components of the soil which give rise to microbial microsites and diffusion gradients which are important determinants of soil function.