Field experiments were conducted from 1979 to 1989 to determine the influence of conventional, reduced, and no-tillage systems and different herbicide combinations on weed species and population, weed control, and soybean injury, population, and yield. In no-till (NT) non-treated plots, there was an abrupt shift from horseweed as the dominant early spring emerging weed to gray goldenrod in 1985. Following its initial observation, gray goldenrod became the dominant species within 2 yr, with giant foxtail as the only other species observed in these plots. Giant foxtail was the dominant weed species from 1980 to 1989 in conventional till (CT) and reduced-till (RT) plots. There also was a shift in the frequency of occurrence and in density of several broadleaf weed species during the 11-yr study. Most herbicides provided excellent control of all weeds in all tillage systems, especially those that included POST herbicides. There was little difference between glyphosate and paraquat in controlling weeds present at the time of planting in NT. PRE herbicides caused 2 to 9% soybean injury with slightly greater injury occurring in CT and RT than in NT. The POST broadleaf herbicides did not significantly increase soybean injury. There were no differences in soybean population or yield among the herbicide treatments regardless of tillage. There also was no difference in soybean population or yield in NT compared with CT when averaged over all herbicide treatments.