During crop domestication plant characteristics have been artificially selected for, with potential unwanted side-effects as some characteristics might have been altered unintentionally. Since productivity in agriculture relies on the interactions between plant and soil, it is a necessity to study the relationship between plant traits, decomposition, microbial community composition and the possible effects on future plant growth.This thesis presents results from several experiments studying how plants influence the soil, through decomposition of plant residues, the soil microbial community assemblage, and its possible consequence for subsequent plant growth. In a crop rotation experiment plant-soil feedback effects have been studied at field scale, with legacy effects of winter cover crops as the primary focus. Influences of plant traits on litter decomposition and microbial community were studied both in a field context and under controlled greenhouse conditions. By comparing the legacy effects of crops with close-relatives from natural grasslands the effects of domestication on litter decomposition and rhizosphere microbial community composition were explored.In the field experiment, it was observed that mixtures of cover crops can perform better than the sum of their parts when the plants in the mixture complemented each other during their growth. Productivity and quality of cover crops was found to promote growth of subsequent main crops, in part through stimulation of soil fungal biomass and feedback effects of decomposing litter. In the greenhouse, growing plants were observed to suppress decomposition of root and shoot litter to varying extent depending on which plant was present and on the quality of the decomposing litter. The results also indicate that domestication has affected plant functional traits in a variety of ways, rather than having predictable effects across a range of crops. Plant functional traits are a useful approach to study legacy effects, as they predict decomposability of plant residues and partially explain the microbial community composition in the rhizosphere. Significance of plant traits as predictors varied with environmental conditions, thus interpretation of the results should be related to its context. This thesis contributes to the understanding of plant-soil interactions, with emphasis on differences and similarities of agricultural and natural ecosystems.