A case study of the seasonal variations in indicator bacteria populations found in water supply wells in a small southern African village has yielded some disturbing findings. This village is typical of the region but has also benefited from the installation of a an experimental, higher- yielding type of wells, a 'collector' or Rannay-type well. In addition to improved yields, it had been assumed that these types of wells would also yield better quality water because of their superior construction. A two- year monitoring programme was carried out to try and understand the seasonal fluctuations in bacteriological quality terms of the hydrological controls. The study has confirmed that water quality (as determined by E.coli population density) in traditional, hand dug well is very variable. Bacterial population density fluctuates throughout the year but rarely drops to zero. This is also true for the collector well, demonstrating that even screened boreholes are pulling water from throughout the regolith, where soil- and ground-water fluxes ensure thorough mixing, and not preferentially from the fissured hardrock as usually supposed. Descriptions of the field techniques employed and resulting data on microbiological E.coli populations, selected hydrochemical determinands and basic hydrological parameters are presented. Analysis and interpretation of the data, together with hypotheses to account for the variability, form a substantial part of the thesis. The implications of the findings on health and general socio-economic development issues, including long-term water supply in similar rural areas, are discussed.