Studying the genetic regulation of carbon utilization in bacteria and yeast, laid the foundation for our current understanding of gene regulation in prokaryotes and eukaryotes. These simple paradigms continue to provide insights into the mechanisms of evolution of gene regulation, that are otherwise difficult to garner. I shall focus on the genetic regulation of galactose utilization in S.cerevisiae and K.lactis, to highlight the role of gene duplication and sub-functionalisation during evolution. Thus, analysis of the mechanisms and regulatory features of galactose genetic switch of these two yeast species allows us to take a close look at the molecular details of evolutionary process. In K.lactis, galactose is sensed by the GAL1 (KlGAL1) encoded galactokinase, a bifunctional protein (this protein also catalyses the phosphorylation of galactose, the first step in the catabolism of galactose). In S. cerevisiae, the signal is mainly sensed by a dedicated signal transducer encoded by GAL3 (it has only signal transduction function but lacks galactokinase activity), an orthologue of KlGAL1 and a paralogue of S.cerevisiae GAL1 (ScGAL1). In addition to GAL3, ScGAL1 encoded galactokinase (bifunctional) also performs the redundant signal transduction function, if cells lack GAL3. Upon interacting with galactose, Gal3/Gal1 protein(s) attains the active conformation. The active Gal3/Gal1 protein(s) sequesters Gal80 protein, the repressor of the switch. Inactivation of Gal80 protein allows the DNA binding transcriptional activator, Gal4 protein, to activate transcription of the GAL genes. I shall focus on the following two questions: why does S. cerevisiae have two signal transducers (ScGAL1 and GAL3) as opposed to a single signal transducer (KlGAL1) in K.lactis? A clue to the above question comes from the unique epistatic interaction between different alleles of GAL80 and GAL4, which I shall discuss in some detail. The second question that I would like to address is the role of stochastic transition between the OFF and the ON state of the GAL switch in S.cerevisiae strain lacking GAL3.