Concentration reduction theory is the leading theory regarding the mechanism of competition for nutrients in soils among plants, yet it has not been rigorously tested. Here we used a spatially explicit, fine-scale grid-based model that simulated diffusion and plant uptake of nutrients by plants in soil to test whether concentration reduction theory was appropriate for terrestrial plant competition for nutrients. In the absence of competition, increasing the rate of diffusion allows a plant to maintain positive growth rates below the lowest average concentration to which it can reduce nutrients in soil solution (R*). As such, differences among plants in the reduction of soil moisture, which here primarily affects nutrient diffusion, can cause R* to predict competitive success incorrectly. The stronger competitor for nutrients captures the largest proportion of the nutrient supply by ensuring nutrients contact its roots before those of a competitor. Although the metric derived from concentration reduction theory, R*, might have predictive power for competitive outcomes in terrestrial ecosystems, this evidence suggests that plants outcompete other plants for nutrients by pre-empting the supply, not reducing the average concentration.