Four month-long field experiments investigated movements of water and solutes through unsaturated sand plains near Princeton, Minnesota, and Oakes, North Dakota. Atrazine and bromide were applied to bare soils and soils planted with corn. The field plots were irrigated according to local farming practices. At the end of each experiment, unsaturated soils were analyzed for atrazine and bromide concentrations and oxygen and hydrogen isotope compositions of soil water. Most soil water was affected by evaporation but groundwater beneath the plots had no evaporative isotopic signature. Therefore most recharge consisted of water that was unaffected by evaporation. Sources of such water may have included snowmelt, prolonged or high-intensity rainfalls that were not interrupted by periods of drying, and water that moved through preferential flow paths. Preferential flow also was suggested by the detection of atrazine, deethylatrazine, and bromide in groundwater shortly after each application of irrigation water at Princeton and by isolated concentrations of atrazine and bromide in soil well below the main masses of chemicals at Oakes.