Habitat succession is thought to influence the importance of competition in assemblages. Competitive interactions are considered of critical importance in structuring ant assemblages, but field experiments show varied effects. I tested how succession in managed boreal forests affects the outcome of competition from dominant red wood ants, Formica aquilonia, through a removal experiment in replicated stands of three different ages (0-4, 30-40, and 80-100 years old). F. aquilonia abundance was reduced by 87%, and procedural controls showed no nontarget effects. The succession gradient revealed the full range of possible responses from ant species: decreases in 1-4-year-old stands, increases in 30-40-year-old stands, and no effects in 80-100-year-old stands, where diversity was lowest. Habitat succession thus regulates competitive interactions in this system. I propose a model for this system, where competitive effects depend on time since disturbance. In this case, soon after disturbance the dominant species facilitates increases in the abundance of other species. At intermediate times, competition reduces the abundance of some species. Finally, in long-undisturbed habitats, competitors may fail to evolve, particularly in high-stress environments. Interactions between competition and habitat succession may explain why structuring effects of ecologically dominant species appear inconsistent.