Hyaluronic acid (HA) is a high-value polysaccharide with many biomedical applications. Microbial production of HA is now replacing the traditional extraction method from rooster combs. Production of medical-grade HA with defined characteristics requires controlled process conditions because there are many fermentation process parameters that affect the microbial synthesis of HA. This necessitates the development of online tools for monitoring multiple analytes during microbial fermentation. Here, we describe the application of in situ transflectance spectroscopy for online quantification of seven major fermentation analytes, viz. biomass, glucose, lactate, formate, ethanol, acetate and HA in metabolically engineered Lactococcus lactis fermentations. The near-infrared spectral information acquired from synthetic mixtures and untransformed L. lactis fermentations were used to develop chemometric models. Based on principal component analysis and partial least squares regression methods, analyte-specific models were developed for quantification. These models were then independently validated for fermentation analytes from four different recombinant L. lactis strains. The chemometric model developed for HA based on recombinant L. lactis fermentation data and pure HA standards could accurately predict HA concentrations under homolactic conditions. The online estimation of HA was found to be poor under heterolactic conditions due to the overlapping absorbance of acetate produced in these cultures. Alternatively, an independent model based on yield correlation was successfully developed for indirect real-time quantification of HA.