The impact of septic tank effluent disposal on groundwater quality was investigated at Oxford. The Oxford township can be regarded as typical of many small communities on the Canterbury Plains which have a high density of septic systems serviced by soakage pit drainage. The primary concern with grouped septic systems is the potential for both chemical and microbial groundwater contamination. The alluvial gravel aquifers of the Canterbury Plains are especially susceptible to microbial contamination due to the high rates of groundwater flow which may transport both bacteria and viral contaminants over large distances. Geological investigations established the presence of an areally extensive, tuff derived, clay unit which forms an aquitard beneath the unconfined aquifer in the north of the Oxford area. Recharge of the unconfined aquifer above the clay unit is exclusively from rainfall infiltration while to the south, groundwater levels respond to rainfall infiltration and influent seepage from the Eyre River. The presence of two hydrogeologically distinct gravel units within the unconfined aquifer was determined by application of the column dilution technique. Point dilution tests showed the average groundwater velocity of 130 m/day in the upper gravel unit to be significantly higher than the 40 m/day measured in the underlying gravels. The presence of discrete channels of preferred flow within the unconfined aquifer system was also established by point dilution tests, intrachannel velocities ranging from 210 to 400 m/day. A resistivity salt tracing test indicated groundwater flow in an easterly direction with a velocity of 250 m/day through an observed channel feature. Groundwater quality monitoring showed a significant degree of groundwater contamination close to the Oxford township. Concentrations of faecal coliform bacteria in excess of drinking water standards were detected up to 900 m downgradient of the nearest septic tank. Elevated levels of chemical indicators (N03-N, CL-) were also detected in all monitoring wells. The pattern of groundwater contamination was complex, reflecting both the heterogeneity of groundwater flow through the unconfined aquifer system and the influence of monitoring well location. Predictive modelling indicated the potential for the transport of faecal coliform bacteria up to 2.6 km downgradient of Oxford. Modelling also suggested increased urban development within Oxford to have a relatively minor effect on the overall extent of groundwater contamination. Additional hydrogeological and water quality data, aided by the application of numerical solute transport modelling techniques, may provide a more accurate estimate of the impact of septic tank effluent disposal on groundwater quality. Future sewage disposal options for Oxford have to balance the low potential for microbial contamination of drinking water supplies outside the groundwater zone delineated by this study, against the environmental acceptability continuing contamination of this zone. This study has identified the need for further research into the effects of septic tank effluent discharge on groundwater quality in the Canterbury region, to provide a sound base for future resource management decisions.