The addition of dicamba as a weed control option in soybean [Glycine max (L.) Merr.] is a valuable tool. However, this technology must be utilized with other herbicide sites of action (SOAs) to reduce selection pressure on weed communities and ensure its prolonged usefulness. A long-term trial was conducted for 7 yr in Indiana to evaluate weed community densities and species richness with four levels of dicamba selection pressure in a corn (Zea mays L.)–soybean rotation. Monocot densities and richness increased over time in the dicamba-reliant treatment. Dicot densities in the dicamba-reliant treatment declined over time, but dicot richness increased. The soil weed seedbank was affected by the varying herbicide strategies. The dicamba-reliant strategy had greater than 43% higher total weed density than all other treatments, primarily due to having a monocot density that was at least 71% higher than the other treatments. The fully diversified strategy with eight SOAs and residual herbicides used every year had the lowest total weed species richness in the soil seedbank, which supported the in-field observations.