The trehalose-P synthase was purified to near homogeneity from the cytoplasmic fraction of Mycobacterium smegmatis. At the final stage of purification, the enzyme preparation showed one major band of 59 kDa on SDS gels. The 59 kDa band became labeled with N3-UDP[32P]-glucose, and this labeling was inhibited in a concentration-dependent manner by either unlabeled UDP-glucose or GDP-glucose. The native enzyme also had a molecular weight of about 60 kDa by gel filtration, indicating that the active enzyme is a monomer. The 59 kDa protein was subjected to endoproteinase Lys-C digestion, and three peptides isolated by HPLC were sequenced. The sequences of 56 amino acids in these three peptides showed 60% identity to the trehalose-P synthases of Saccharomyces cerevesiae and Schizosaccharomyces pombe. The purified mycobacterial enzyme catalyzed the synthesis of trehalose-P from glucose-6-P and a variety of nucleoside diphosphate glucose derivatives, depending on whether a polyanion was absent or present. Thus, UDP-glucose and GDP-glucose were the best glucosyl donors, but maximum activity with UDP-glucose required the presence of a polyanion such as heparin, whereas activity with GDP-glucose was relatively independent of polyanion. The presence of heparin in the incubation mixture increased the affinity of the enzyme for UDP-glucose by a factor of 100, or more. However, the affinity for GDP-glucose was only twofold better in the presence of heparin. The purified synthase also utilized ADP-glucose and CDP-glucose, but the Km for these glucosyl donors was quite high even in the presence of polyanion. The effect of heparin on UDP-glucose activity was dose-dependent and maximum at about 1–2 µ;g of heparin/incubation. However, the size of the heparin molecule (i.e., the number of monosaccharide residues) was critical for activation, and only those heparins with 18 or more monosaccharide units were effective in stimulating activity.