<p>The occurrence of highly exothermic and potential runaway reactions represents a significant hazard to workers, facilities, and the surrounding ecosystem. Due to their high reaction velocity, these reactions often prove difficult to monitor. In many instances, the recorded data result from several single-quantity sensors, such as temperature and pressure sensors. Yet, deviations from the anticipated process, and thus potentially hazardous reaction parameters, may only be identified by the combination of the sensor data, leaving the cause of the deviations unexplained. In this study, a custom-built microreactor in combination with in-line near-infrared spectroscopy and off-line gas chromatography was employed to allow monitoring of the acid-catalyzed esterification of acetic acid and methanol. Using a near-infrared spectrometer in conjunction with partial least squares regression models, reactant and product concentrations could be determined at various residence times and temperatures. This approach enabled the safe and efficient investigation of previously unexamined parameter ranges and the expansion of the design space of experiments. The near-infrared spectra were subjected to principal component analysis classification in order to obtain additional qualitative information. The conversions and space–time yields of the microreactor could be calculated from the obtained data. The values indicated that the microreactor design delivered yields up to twice as high as those of commonly used batch reactors. This study hence demonstrates the efficacy of spectroscopic methods for reaction monitoring in microreactors and their potential for application in micro process analytical technologies.</p> Graphical Abstract <p></p>