Near-infrared spectroscopy (NIR) was implemented in the resin transfer molding (RTM) process to inline monitor the degree of curing of a bio-based epoxy resin, which consists of epoxidized linseed oil (resin) and citric acid (hardener), respectively. A NIR micro-spectrometer was used for the development of robust calibration models using partial least squares (PLS) regression. Since the micro-spectrometer offers a smaller wavelength range compared with conventional NIR devices, and typical absorbance peaks are not directly involved in the captured data range, the results show new insights for the utilization of this technology. Different pre-treatments of the spectroscopic data have been tested, starting with different reference spectra, i.e., uncured resin and polytetrafluorethylene (PTFE), and followed by chemometrical algorithms. As a reference method for the degree of curing, direct current (DC) supported by differential scanning calorimetry (DSC) was used. The results show the potential of these cost-efficient and compact NIR micro-spectrometers for the intended inline monitoring purpose to gain relevant information feedback during the process.