1. Branching enzymes from rat and rabbit liver, as well as from potato and maize were prepared. They were almost free from contaminating glucan-degrading enzymes.2. In 'sweet corn' maize, two separate fractions with (al,4)glucan : (al,4)glucan a6-glycosyltransferase activities were obtained. One of them synthesized amylopectin, the branched component of starch, in the presence of phosphorylase and Glcl P, while the other fraction synthesized phytoglycogen. Furthermore, in a maize variety which does not accumulate phytoglycogen, only one fraction of branching activity was found, that formed amylopectin under the above-mentioned conditions.3. Comparative analyses performed with native (al,4)-(a1,6)glucopolysaccharides, and those synthesized in vitro with the branching enzyme from the same tissue, demonstrated a close similarity between both glucans.4. It may be concluded that the branching enzyme is responsible for the specific degree of (a1,6) branch linkages found in the native polysaccharide.The branching enzyme transfers to an acceptor an (al,4)-linked glucan fragment removed from a donor by breakage of an (a1,4) bond, which may or may not be the original donor, creating an (a1,6) bond [l, 21. The reaction does not involve net glucan synthesis.The enzyme responsible for the formation of (a1,4) glucosidic bonds in vivo is the synthase, nevertheless it is also possible to create these bonds through the reverse action of phosphorylase. Much valuable information on the regulatory mechanism of glycogen synthase or phosphorylase has been published [3] but little has been reported on branching enzymes. Thus, it has been assumed that a relatively high elongation activity [synthesis of (al,4)glucosidic chains] would result in the synthesis of glycogen with a relatively low degree of branching. On the other hand, a relatively low elongation activity would allow the branching enzyme to introduce a high proportion of (al,6)glucosidic linkages [3].Polysaccharides with different numbers of (al,6)glucosidic linkages are normally found in nature; so starch amylopectins from different origins are different. Potato amylopectin is quite different from that of maize or wheat. Also, rat liver glycogen is quite different from rabbit liver glycogen or bacterial glycogen. Thus, we wondered whether the density of branching points depends on the activity of the branching enzyme, on the ratio of branching to synthase activities or if it is due to intrinsic characteristics of the branching enzyme activity.The numbers of branching points of native (a1,4)-(al,6)glucopolysaccharides were quantified and compared to those of the polysaccharides synthesized in vitro. The branching enzymes were obtained from the respective tissues and the products compared by a method developed in our laboratory [4].The results reported here show that the degree of branching is related to intrinsic properties of the branching enzymes, rather than to the ratio of elongation/branching activities.