Field experiments were conducted to determine optimum nozzle settings for applying glyphosate [N-(phosphonomethyl)glycine] in the recirculating sprayer for postemergence control of johnsongrass [Sorghum halepense (L.) Pers.] in soybeans [Glycine max (L.) Merr.]. Herbicide sprays were directed across the row to johnsongrass growing taller than soybeans in July and August. Herbicide not sprayed on johnsongrass was trapped and reused. Glyphosate at 0.56, 1.12, and 2.24 kg/ha applied with commercially available 25° spray nozzles provided johnsongrass control and soybean yields equal to those following applications with specialized uniform droplet nozzles. Glyphosate at 1.7 kg/ha applied in the recirculating sprayer using only one nozzle per row provided control of johnsongrass equal to or better than that from applications made with two, three, or four nozzles per row. Soybean yield following application of glyphosate at 1.7 kg/ha with one nozzle per row was equal to yields obtained following its application with two, three, or four nozzles per row, with or without surfactant at 0.1% in spray solutions. Soybean yield was higher with four nozzles per row than with one nozzle per row when 0.5% surfactant was included in spray solutions. Soybean injury was lower and yield was higher when glyphosate was applied in the recirculating sprayer rather than over-the-top with a conventional sprayer. Glyphosate at 1.12 kg/ha applied in the recirculating sprayer caused more injury to ‘Hill’ and ‘Bragg’ than to ‘Forrest’ or ‘Tracy’ soybeans.