Conservation tillage systems, such as no-tillage, are ecologically advantageous because they reduce soil erosion; however, they rely heavily on herbicide use. Our goal was to determine how weed communities of no-tillage systems are affected when the system is modified to reduce herbicide use through a combination of banded herbicides and interrow cultivation. To this end, we conducted a 9-yr study in a no-tillage corn–soybean–winter wheat rotation. All management systems had a preplant application of glyphosate, followed by either broadcast PRE herbicides (conventional no-tillage), interrow cultivation with banded PRE herbicides, or interrow cultivation alone. Aboveground weed densities were assessed each year and data were grouped into early (1991 to 1993) and late (1996 to 1998) time periods. Over time, weed communities became more distinct, showing a strong response to management and crop. In the early years, weed communities separated more in response to management than crop. In the late years, this was reversed. Weed communities in systems with interrow cultivation were more diverse than those in conventional no-tillage. The response to weed management system and crop was species specific. For example, the abundance of yellow foxtail was higher when interrow cultivation was employed, but abundance was equal in all crops. Dandelion was more abundant in conventional no-tillage of corn and soybean; however, it was equally abundant in all management systems in wheat. Seed bank species richness increased over time and was highest in systems with interrow cultivation. Herbicide use can be reduced in a modified no-tillage corn–soybean–wheat rotation by incorporating interrow cultivation, with or without banded herbicides, into the management plan. The weed community trajectory changes, and the weed community becomes more diverse. A more diverse weed community will not necessarily alter how we manage weeds.