The alluvial aquifer within a 160-acre area of the Rio Grande de Manati alluvial plain was investigated to evaluate its potential as a water-supply source for the Barrios Rio Arriba Saliente and Pugnado Afuera, municipio of Manati, Puerto Rico. Analysis of well boring samples and the results of electric resistivity surveys indicate that the average thickness of the unconsolidated alluvial deposits in the study area is about 100 to 110 feet. The alluvium is a mixture of sand and gravel, which generally has a porosity of 0.2 to 0.35. Short-duration pump tests in small-diameter piezometers indicate that the alluvial aquifer has a hydraulic conductivity of about 200 feet per day and a transmissivity of about 7,900 feet squared per day. Analyses of water levels in piezometers, combined with stage measurements at a series of surveyed reference points along the Rio Grande de Manati channel, indicate that the water-table gradient in the alluvial aquifer is about 0.001, and that ground-water flow is generally from south to north, in the general direction of river flow. The water-table data indicate that the Rio Grande de Manati is the principal source of ground-water recharge to the alluvial aquifer in the study area. Because base flow for the Rio Grande de Manati is usually greater than 44 cubic feet per second, a continuous withdrawal rate of 0.5 to 1.0 cubic foot per second (225 to 450 gallons per minute) from a production well is possible. Chemical analysis of a ground-water sample indicates that the alluvial aquifer water meets U.S. Environmental Protection Agency secondary standards for selected constituents. Bacteriological analysis of ground-water samples indicates that the ground water contains little or no fecal coliform or fecal streptococcus bacteria. Although long-term data from upstream of the study area indicate high levels of fecal coliform and fecal streptococcus prior to 1996, bacteriological analyses of Rio Grande de Manati water samples obtained during the present study indicate that fecal coliform and fecal streptococcus concentrations are within the standards for surface water intended for use (or with the potential for use) as a raw source of public water supply in Puerto Rico. If a production well were constructed in the study area, it would be located close to the river channel (within 500 to 800 feet). Pumping from the porous and permeable alluvial aquifer close to the river channel could substantially enhance recharge from the Rio Grande de Manati channel to the aquifer. Enhanced recharge could shorten travel times for ground water in the aquifer, which might not allow sufficient time to attenuate bacteria and viruses. Travel times for bacteria moving from the river channel to a hypothetical production well were estimated using the numerical transport model MODFLOW/MT3DMS with an uncalibrated model of the alluvial aquifer. The model assumes a well pumping at 1 cubic foot per second. The transport of particles from the river to the well is most sensitive to the porosity of the aquifer and the pumping rate of the well. Sensitivity analysis indicates that a decrease in pumpage will increase the time of travel for particles to move from the river to the pumping well. The model indicates that the leading edge of a plume would reach the production well in about 40 days assuming a porosity of 0.20, 60 days assuming a porosity of 0.275, and about 70 days assuming a porosity of 0.35. If the well were moved 50 feet further from the river, the leading edge of the plume would reach the well in about 50 days assuming a porosity of 0.20 and about 70 days assuming a porosity of 0.275. These estimates are considered worse case estimates because no decay rate was included in the simulation, and because the hypothetical well was located in the center of the alluvial plain rather than further eastward, away from the river channel.