Roots are a vitally important plant characteristic but they are difficult to study and evaluate in soil or even in artificial media. The objective of this research was to determine if an x-ray based system could be used to rapidly and accurately evaluate and classify soybean (Glycine max) roots. We chose a normal rooting variety, 'USG 5601T' and a prolific rooting accession, PI416.937 to utilize in the study. Ten seedlings of each variety were transplanted into artificial substrate containers and imaged with a digital x-ray radiography system at transplanting and at 7 d intervals through day 21. The x-ray images were analyzed manually to determine if individual plants could be classified into variety-type classes of root architecture. At 14 d post-transplant, the individual plants could be placed into USG 5601T or PI clusters with approximately 80% accuracy. This was based on taproot length and lateral root density values. Additional data will be presented from plants grown in larger substrate containers and imaged weekly up to 42 d.