Microbial mannanases are biotechnologically important enzymes since they target the hydrolysis of hemicellulosic polysaccharides of softwood biomass into simple molecules like manno-oligosaccharides and mannose. In this study, we have implemented a strategy of molecular engineering in the yeast Yarrowia lipolytica to improve the specific activity of two fungal endo-mannanases, PaMan5A and PaMan26A, which belong to the glycoside hydrolase (GH) families GH5 and GH26, respectively. Following random mutagenesis and two steps of high-throughput enzymatic screening, we identified several PaMan5A and PaMan26A mutants that displayed improved kinetic constants for the hydrolysis of galactomannan. Examination of the three-dimensional structures of PaMan5A and PaMan26A revealed which of the mutated residues are potentially important for enzyme function. Among them, the PaMan5A-G311S single mutant, which displayed an impressive 8.2-fold increase in k(cat)/K-M due to a significant decrease of K-M, is located within the core of the enzyme. The PaMan5A-K139R/Y223H double mutant revealed modification of hydrolysis products probably in relation to an amino-acid substitution located nearby one of the positive subsites. The PaMan26A-P140L/D416G double mutant yielded a 30% increase in k(cat)/K-M compared to the parental enzyme. It displayed a mutation in the linker region (P140L) that may confer more flexibility to the linker and another mutation (D416G) located at the entrance of the catalytic cleft that may promote the entrance of the substrate into the active site. Taken together, these results show that the directed evolution strategy implemented in this study was very pertinent since a straightforward round of random mutagenesis yielded significantly improved variants, in terms of catalytic efiiciency (k(cat)/K-M).