[ACCESS RESTRICTED TO THE UNIVERSITY OF MISSOURI-COLUMBIA AT REQUEST OF AUTHOR.] Soybean (Glycine max (L.) is currently grown throughout the world because it has been adapted to many environments and because of the high protein and oil content of the seeds. Water scarcity is responsible for the biggest crop losses worldwide and this is expected to worsen; thus, much attention is directed towards the development of drought tolerant crops. The root system is fundamentally important for plant growth and survival because of its role in water and nutrient uptake. Crops with deep roots can capture more soil resources, particularly water, to support shoot growth and yield formation. However, the investigation of root systems is difficult and remains challenging, especially under field conditions. Nonetheless, a better understanding of root system form and function is critical to develop strategies to breed for more stress-resilient crops for local production environments. Studies of soybean root systems in general, and rooting depth in particular have been limited. Thus, the aims of the research described in this dissertation were to (i) identify genotypic diversity in rooting depth and distribution of roots in the soil profile and relate these traits to above ground characteristics including yield under rainfed field conditions in a wide range of soybean genotypes, (ii) characterize, compare and contrast root systems of selected soybean genotypes grown under field- and greenhouse-conditions, and (iii) explore the influence of scion and rootstock genotype on root growth of contrasting soybean genotypes under well-watered and water deficit stress conditions. In the first series of experiments, a set of five soybean genotypes that represented contrasting root rooting depths and root elongation rates were selected based on greenhouse experiment and grown under rainfed field conditions. The core break method was used to assess root distributions of these genotypes in two years. The main goals of this experiment were to confirm genotypic variation for key root traits, including rooting depth and distribution, and to determine whether rooting depth is related to seed yield and selected shoot traits. This study confirmed significant variation among genotypes regarding their rooting depth and root distribution in the soil profile. Genotypes with greater maximum rooting depth also exhibited greater numbers of roots in the lower soil strata than shallower rooting genotypes, and rooting depth was positively correlated with seed yield. Confirmation of differences in rooting depth among these genotypes and the relationship with seed yield under field conditions establishes the suitability of the selected genotypes for physiological studies, studies of genetic mechanisms underpinning maximum rooting depth in soybean, and to confirm the potential for yield increase as a result of selection for deep rooting. A second study consisted of two greenhouse experiments to evaluate the effect of water availability on the rooting depth plasticity of deep- and shallow-rooted genotypes. Six contrasting genotypes were grown in PVC pipes under well-watered and dry-down conditions. The soil media was a mixture of soil and sand with a ratio of 4:1, respectively. Significant genotype, water treatment, and genotype by water treatment interaction effects were observed for maximum rooting depth. Maximum rooting depth increased in the dry-down compared to the well-watered treatment and induced a reallocation of root length from shallow strata to deeper regions in the profile for all genotypes. The extent of the difference in rooting depth between well-watered and dry-down treatments, measured as plasticity, was significantly different among genotypes. Thus, plasticity in maximum rooting depth appears to be under genetic control in soybean and may be a suitable target for breeding efforts aimed at increasing yields under drought. In a final study, the influence of scion and rootstock genotype on shoot growth and root system characteristics was examined in deep tubes in an automated rainout shelter. Plants were sown into 1.5- m deep tubes filled with a soil-sand mix (4:1) and grown under well-watered and dry-down conditions. Nine days after sowing, self and reciprocal grafts were made using the wedge grafting method. The dry-down treatment resulted in significantly increased rooting depth for all grafted as well as the non-grafted treatments compared to well-watered treatment. As expected, root length densities in the top 30 cm of the soil were greater for well-watered plants than plants in the dry-down treatment whereas the opposite was true for root length density at depth. Overall, whether self-grafted or serving as rootstock only, the deep-rooted genotype had a stimulatory effect on root growth in most soil strata, particularly under dry-down conditions. In general, limited differences observed among the grafting treatments suggest a small influence of the scion or rootstock genotype on the rooting depth and root distribution in the soil profile. However, grafting studies with additional genotypes should be conducted to explore whether this observation is specific to the genotype combination used in this study or whether it applies more generally for soybean. The experiments described in this dissertation lay the foundation for additional physiological and genetic studies. Further research is needed to ascertain the physiological mechanism behind the responses of contrasting genotypes, and to identify molecular markers and/or genes to facilitate incorporation of desirable root traits into a breeding program to increase yields and/or yield stability under drought conditions.