Instances of glyphosate resistance in weeds are increasing in Iowa. Farmers may face significant economic loss from improperly managing glyphosate resistance of some weed species in soybean (Glycine max (L.) Merr.) production. The objective of this study was to utilize yield monitoring technology and digital aerial imagery to compare soybean yield and economic performance of the traditional weed management based on two applications of glyphosate (G2)  with a program replacing first glyphosate application with a residual herbicide (R+G), Extreme. The comparisons were made in at least three alternating strips in eight field-scale on-farm trials in 2008 and in 20 trials in 2009. Both treatments received a second glyphosate application in 15-50 days after the first herbicide application. Late-season digital aerial imagery was used to evaluate soybean canopy differences and characterize spatial variability in yield differences (YD) calculated between R+G and G2 treatments every 30 to 50 m along the herbicide strips. Hierarchical analysis was used to identify field and within-field level factors that affected the observed YD. On average in each year, R+G treatments produced a slightly reduced soybean yield (by 15 or 20 kg ha^-1) than G2 treatments. The additional cost of residual herbicide was about $45-50 ha^-1 more than that of the two applications of glyphosate. Trials receiving >50 cm of cumulative March through May rainfall in 2008 were about 3 times more likely to have an economic soybean yield loss (>100 kg ha^-1) in strips where Extreme herbicide was used than those trials receiving <50 cm of early season rainfall. On-farm studies using precision agriculture technologies should help find alternative weed management systems that rely less on glyphosate and quantify the potential economic loss or benefits from reduced glyphosate applications on soybean.