Both abiotic and biotic gradients exist in soils, and several of these gradients have been shown to select for plant traits. Moreover, plants possess a multitude of traits that can lead to strong niche construction (i.e. plant-induced changes to soils). Our objectives in this paper are to outline both empirical and theoretical evidence for the evolutionary consequences of plant-soil linkages and feedbacks on plants along soil heterogeneity gradients. We describe a simple mathematical model of plant evolution to explore the relationship between the sign and magnitude of feedback and the divergence of plant traits. We also constructed an individual-based simulation model to study the conditions under which plant-soil feedbacks occur, niche construction evolves, and plant traits diverge. This approach allows us to address specific hypotheses regarding relationships between positive and negative plant-soil feedback with variation in niche construction, the strength of selective gradients and the relative importance of local adaptation vs. feedbacks. The models suggest that feedbacks between soils and plants may commonly result in evolutionary interactions. The simulation model indicates that plant traits can diverge with niche construction and traits can be selected for in response to niche construction. However, the magnitude of feedbacks and how strongly they evolve depends on the amount of gene flow and the strength of selective gradients over time. These results suggest that plant-soil feedback can lead to evolution in plants and reveals new research directions for further inquiry. Questions addressing trade-offs and relationships between positive and negative feedbacks as well as adaptation and maladaptation of plant traits represent important frontiers in plant-soil feedback studies.