We report here that the open reading frame YKL248, previously identified during the systematic sequencing of yeast chromosome XI [Purnelle B., Skala, J., Van Dijck, L. & Goffeau, A. (1992) Yeast 8, 977–986] encodes UDP-glucose pyrophosphorylase (UGPase), the enzyme which catalyses the reversible formation of UDP-Glc from glucose 1-phosphate and UTP. Proof for this function come from sequence alignment of the YKL248 product with UGPase of other species, from complementation studies of an Escherichia coli galU mutant deficient in UGPase activity, and from overexpression studies. In particular, the amino acid sequence motifs involved in the binding of glucose 1–phosphate and UDP-Glc are entirely conserved between the yeast, bovine, human and potato tuber UGPases, and multi-copy expression of YKL248 resulted in a 40-fold increase in UGPase activity. This gene was, therefore, renamed UGP1. Gene disruption at the UGP1 locus in a diploid strain, followed by tetrad analysis, showed that UGPase is essential for cell viability. Functional analysis of UGP1 was, therefore, carried out by generating strains in which UGPase could be either overexpressed or depleted. This was done by generating haploid strains carrying either UGP1 on a multicopy vector or the chromosomal deletion of UGP1, and rescued by a vector bearing the wild-type gene under the control of the glucose-repressible galactose-inducible promoter. The effects of overproducing UGPase on the cell metabolism and morphology were carbon-source dependent. On glucose medium, the 40–fold increase of UGPase activity was restricted to a twofold increase in the concentration of glycogen and UDP-Glc, with no significant effect on growth. In contrast, on galactose, the 40–fold increase in UGPase activity was accompanied by several effects, including a threefold reduction of the growth rate, a 3–5–fold increase in the concentrations of UDP-Glc, UDP-Gal and galactose 1–phosphate, a higher sensitivity to calcofluor white and an increase in the degree of protein glycosylation. Depletion of UGPase activity was performed by transferring the mutant strains from galactose to glucose medium. Unexpectedly, growth of these mutants on glucose was as efficient as that of the control, although the mutants contained only 5–10% wild-type UGPase activity, and a growth defect could never been obtained, even after serial transfers of the mutants to a 10% glucose medium. However, the 10–fold reduction of UGPase activity induced a multi-budding pattern, a higher resistance to zymolyase, a slight increase in the calcofluor sensitivity and a decrease in the cell-wall β-glucan content. All these alterations, induced by manipulating the UGP1 gene, are discussed in the context of the strategic position of UDP-Glc in yeast metabolism.