Roots from Grimm alfalfa that had been subjected to different cutting treatments were collected during summer, fall, and winter and the amount of reserves determined;Frequent cutting checked root growth 50 percent or more. The root reserves were higher in the uncut plants than in the cut plants. The period of greatest difference was during October and November;The amount of reserve materials in the roots of cut and uncut plants was positively correlated with cold resistance in both artificial freezing tests and in field tests;The starch and dextrin decreased in autumn and early winter. The non-reducing sugars increased as fall progressed;The soluble nitrogen increased during fall, dropped during winter and increased again in the spring. The colloidal nitrogen remained fairly constant during the winter and spring;The amount of fall top growth of hardy and non-hardy varieties of alfalfa was inversely proportional to hardiness. The hardy varieties made very little top growth in late fall; the tender varieties made extensive fall growth;Tests on the stability of the protoplasm were made and no consistent relation was found between the amount of nitrogen precipitated by freezing a watery extract of alfalfa proteins and cold resistance;Six varieties of alfalfa were analyzed for their content of root reserves. No variety was so different from the others that it could be placed in its proper hardiness group on the basis of reserves. The trend of the root reserves was quite similar in all varieties during fall, winter and spring;A pectic-like constituent was isolated from alfalfa roots, but showed no clear cut relation to cold resistance.