This study investigated the effect of water potential on soil microbial C and N pools. Two soil types were treated with additions of salt solution to establish osmotic water potentials, and by ceramic plate - pressure chamber apparatus to establish matric water potentials. Soils were then subjected to short-term incubations. Soil microbial C and N contents (BC and BN) were measured mainly by the fumigation-extraction and fumigation-incubation methods. Results showed that both Microbial C and N pools were markedly affected by soil water potential. The soil microbial C content always showed an increase with increasing water stress and then a decrease beyond a threshold value of water stress, compared to the microbial C content at a control water potential of -0.03 MPa (-0.3 Bar). This response pattern to water stress was true, not only for osmotic stress, but also for matric stress, and regardless of the osmotic agent employed. The response pattern of the microbial N pool to water stress generally contrasted with that of the C pool, and depended on the osmotic strength of the extraction solution (K₂SO₄) used in the determination. Non-isotonic extraction after fumigation resulted in a decrease in microbial N content with increasing water stress, while isotonic extraction resulted in an increase with increasing water stress, beyond a threshold value of water stress. Soil microbial C/N ratio always increased with increasing water stress. Matric water stress had a more marked effect on BC and BN than osmotic stress. The possible reasons for the response patterns of BC, BN and microbial C/N ratio have been discussed in this thesis. Some suggestions on the methodology of microbial biomass measurement for water stressed soil samples have been made, and mainly relate to the biomass fumigation techniques and possible changes in the Kc, Kce and Kn values under water stress, and to the substrate induced respiration (SIR) method and suppressed respiration under water stress.