Studies were conducted in 1997 and 1998 to evaluate the efficacy and economics of glyphosate-resistant and nontransgenic soybean systems. The three highest yielding glyphosate-resistant and nontransgenic soybean cultivars were chosen each year for three Mississippi locations based on Mississippi Soybean Variety Trials. Treatments within each cultivar/herbicide system included nontreated, low input (one-half of the labeled rate), medium input (labeled rate), and high input level (labeled rate plus an additional postemergence application). In 1997, all systems controlled hemp sesbania by more than 80% but nontransgenic systems controlled hemp sesbania more than the glyphosate-resistant systems in most instances in 1998. High input levels usually controlled pitted morningglory more than low or medium inputs in 1997. In 1998, both systems controlled pitted morningglory by 90% or more at Shelby; however, at other locations control was less than 85%. Soybean yield in 1997 at Shelby was more with the glyphosate-resistant system than with the nontransgenic systems at medium and high input levels, primarily because of early-season injury to a metribuzin-sensitive cultivar in the nontransgenic system. In 1998, soybean yield at Shelby was more with the nontransgenic system than the glyphosate-resistant system, regardless of input level, due to poor late-season hemp sesbania control with glyphosate. Net returns were often more with the glyphosate-resistant system at Shelby in 1997. Within the glyphosate-resistant system, there were no differences in net return between input levels. Within the nontransgenic system, low input level net returns were higher compared to medium and high input levels due to higher soybean yield and less herbicide cost. At Brooksville, using high input levels, the glyphosate-resistant systems net returns were $55.00/ha more than the nontransgenic system. Net returns were higher with the nontransgenic system compared to the glyphosate-resistant system at Shelby in 1998, regardless of input level.