Challenges such as climate change and a growing world population has increased the need for sustainable solutions that can increase crop production. Microbial inoculants have been proposed to bring forth the next green revolution to agriculture as alternatives to chemical fertilizers and pesticides. Alleviation of abiotic stress, pathogen attack, and nutrient deficiency are some of the many plant benefits attributed to microorganisms. Yet, microbial inoculants have had limited success as commercial, effective products, in part due to our insufficient understanding of the many underlying mechanisms controlling how plants, rhizosphere microorganisms and soil interact. We need to further our fundamental understanding of this complex system, not only to attain more effective microbial inoculants for agriculture, but also to ensure that they truly are sustainable alternatives to chemical fertilizer and pesticides without any unintended side-effects. In this thesis, I present work illustrating that abiotic conditions can alter the plant-microbiota interaction. When soil is nitrogen-limited, the soil microbiota induces nitrogen deficiency in plants. I also show that shoot biomass was reduced by 12% regardless of nitrogen availability, when plants were inoculated with soil microbiota. This specific plant-microbiota interaction was deleterious to plant growth even in conditions free of abiotic stress and pathogen infection. It appears that interacting with microorganisms, is not necessarily costfree for plants. Furthermore, I show that artificial selection of microbiotas associated with low shoot greenness also leads to some plants to exhibiting symptoms of iron deficiency. This finding illustrates that although microbiota engineering can successfully be applied to alter a plant trait, we may also in the process inadvertently alter one or several other plant traits. This highlights why it is essential that we further our understanding of how microorganisms affect plants and under which conditions. In the thesis, I also present a new experimental design for studying plant-microbiota-soil interactions under highly controlled conditions. Due to the complexity of the plant-microbiota-soil interaction, we need such systems wherein microbial contamination, from seeds and the surrounding environment, is minimized to decrease stochastic variation, so that we gain a mechanistic understanding of how microorganisms affect plants. My work and results are a small step towards closing the many knowledge gaps that prevent us from predicting the net effect specific microorganisms will have on a specific plant under specific abiotic conditions. It is important not only to focus on the benefits that microorganisms can provide to plants but also to understanding the deleterious interactions. When we fully understand how and when soil microorganisms impede plant growth and health, we may also be able improve plant growth via microorganisms by minimizing the deleterious interactions.