Coagulation of milk is one of the most important steps in cheese manufacture. Cutting the coagulum at optimum firmness is important to optimise the yield and quality of the cheese produced. The aim of this study was to investigate a prototype sensor to monitor rennet-induced coagulation of skim milk at different protein concentrations (3.3%, 4.0% and 4.7%) and to develop a model to predict the coagulum cutting time at a desired storage modulus (G 0). Fluorescence and infrared backscatter profiles were recorded at wavelengths of 350 and 880 nm, respectively. Rheological measurements were used as a reference method to determine the times required for the coagulum to reach G 0 values of 0.5, 5 and 20 Pa. Time parameters extracted from the optical profiles generated during the coagulation process were used to develop a model to predict the cutting time at which the coagulum reaches selected G 0 values. This study demonstrated that the investigated prototype sensor, combined with the developed prediction model, can be used as an in-line PAT tool for real-time monitoring of milk coagulation and prediction of cutting time in cheese manufacturing.