Chiral amines are essential precursors in the production of biologically active compounds, including several important drugs. Among the biocatalytic strategies that have been developed for their synthesis, the use of omega-transaminases (omega-TA) appears as an attractive alternative allowing the stereoselective amination of prochiral ketones. However, the problems associated with narrow substrate specificity, unfavourable reaction equilibrium and expensive amine donors still hamper its industrial application. The search for novel enzymes from nature can contribute to expand the catalytic repertoire of omega-TA and help to circumvent some of these problems. A genome mining approach, based on the work described by Hohne et al., was applied for selection of potential R-omega-TA. Additional criteria were used to select an enzyme that differs from previously described ones. A candidate R-omega-TA from Capronia semiimmersa was selected, cloned and expressed in Escherichia coli. Interestingly, alignment of this enzyme with previously reported TA sequences revealed the presence of two additional amino acid residues in a loop close to the active site. The impact of this change was analysed with a structural model based on crystallized R-omega-TAs. Analysis of the substrate specificity of R-omega-TA from C. semiimmersa indicates that it accepts a diversity of ketones as substrates yielding the corresponding amine with good yields and excellent enantioselectivity. The expressed enzyme accepts isopropylamine as amine donor what makes it suitable for industrial processes.