Short-chain fructooligosaccharides (scFOS) are valuable health-promoting food additives. During the batch production of scFOS from sucrose the beta-fructofuranosidase catalyst is subject to product inhibition by glucose. Engineering the enzyme for reduced sensitivity to glucose could improve product yields or process productivity while preserving the simple industrial batch design. Random mutagenesis is a useful technique for engineering proteins but should be coupled to a relevant high-throughput screen. Such a screen for sucrose and scFOS quantification remains elusive. This work presents the development of a screening method displaying potential high-throughput capacity for the evaluation of beta-fructofuranosidase libraries using Fourier transform mid-infrared attenuated total reflectance (FT-MIR ATR) spectroscopy and multivariate analysis. A calibration model for the quantification of sucrose in enzyme assay samples ranged from 5 to 200 g/l and the standard error of prediction was below 13 g/l. A library of the Aspergillus japonicus fopA gene was generated by error prone PCR and screened in Saccharomyces cerevisiae. Using FT-MIR ATR spectroscopy, potential hits were identified as those variants that converted more sucrose in the presence of the glucose inhibitor than the parent. Subsequent analysis of reaction products generated by top performers using high-performance liquid chromatography identified a variant producing higher scFOS levels than the parent. At the peak difference in performance the variant produced 28 % more scFOS from the same amount of sucrose. This study highlights the application of FT-MIR ATR spectroscopy to a variant discovery pipeline in the directed evolution of a beta-fructofuranosidase for enhanced scFOS production.