Riparian municipal wells, that are located on riverbanks, are specifically designed to capture a portion of the river water through induced infiltration. Runoff from agricultural watersheds is found to carry enormous amounts of pesticides and nitrate. While the risk of contamination for a vast majority of sites with small-capacity vertical wells is low, potential exists for medium to large capacity collector wells to capture a fraction of the surface water contaminants during flood. Prior monitoring and current modeling results indicate that a small-capacity (peak pumpage 0.0315 m3/s) vertical bank filtration well may not be affected by river water nitrate and atrazine even during flood periods. For a medium capacity (0.0875-0.175 m3/s) hypothetical collector well at the same site, potential exists for a portion of the river water nitrate and atrazine to enter the well during flood periods. Various combinations of hydraulic conductivity of the riverbed or bank material were used. For nitrate, it was assumed either no denitrification occurred during the period of simulation or a half-life of 2 years. Equilibrium controlled sorption (organic carbon partition coefficient of 52 ml/g) and a half-life of between 7.5 and 15 weeks were considered for atrazine. Combinations of these parameters were used in various simulations. Peak concentrations of atrazine or nitrate in pumped water could vary from less than 1% to as high as 90% of that in the river. It was found that a combination of river stage, pumping rates, hydraulic properties of the riverbed and bank, and soil/pesticide properties could affect contaminant entry from river water to any of these wells. If the hydraulic conductivity of the bed and bank material were low, atrazine would not reach the pumping well with or without sorption and degradation. However, for moderately low permeable bank and bed materials, some atrazine from river water could enter a hypothetical collector well while pumping at 0.0875 m3/s. It was interesting to note that doubling the pumpage of this collector well would bring in more ground water from the aquifer (with no atrazine) and thus have a lower concentration of atrazine in the filtrate. For highly conductive banks, it is possible to find some atrazine at a vertical well for a sustained pumpage rate of 0.0125 m3/s if the effect of sorption is neglected. However, with equilibrium sorption, the concentration would be below the detection limit. On the other hand, if a collector well of capacity 0.0875 m3/s is used at the place of the vertical well with highly conductive banks, atrazine concentration in the filtrate would be about 80% of river water even assuming equilibrium sorption and a half-life of 7.5 weeks. Remediation of river water contamination of the aquifer using 'scavenger' wells between the river and the pumping well(s) was not a feasible option due to the contact of the aquifer with a highly conductive bank at the site. However, moving the existing pumping well(s) 100 m upstream would have negligible impact from the bank-stored water. ?? 2002 Elsevier Science B.V. All rights reserved.