This work investigates the use of in situ process analytical technologies, namely, attenuated total reflectance Fourier transform infrared spectroscopy (ATR FTIR) and ATR UV-vis, coupled with a robust chemometric multivariate analysis approach for the determination of the solubility of ketoprofen in organic solvents within the temperature range of 273.15-303.15 K. Principal component analysis of the raw spectral data is applied along with multiple regression analysis and a stepwise regression procedure as the chemometric approach to build inferential models able to predict solubility data. For each solvent, the constructed models are used for solubility determinations and the results are compared with those previously determined by a gravimetric method. Both in situ FTIR and UV-vis render an accurate prediction of solubility data, with a notably high number of data points enabling the subsequent thermodynamic analysis. However, the accuracy of the solubility rendered from the spectroscopic analysis is not as accurate as by the gravimetric method at low temperatures. In addition, the experimental enthalpies of solution are determined by solution calorimetry and are found to be in a range of 18.6-32.4 kJ/mol in the different solvents. The order of solubility data obtained in the solvents studied is discussed in terms of the strength of the interactions with the ketoprofen carbonyl groups analyzed by the C=O frequency of the FTIR data and based on the solution thermodynamic data obtained. All solutions of the work are relatively nonideal, with activity coefficients deviating from unity and enthalpies of mixing deviating from zero. The thermodynamic analysis and results clearly show that calorimetric enthalpies of solution are theoretically and quantitatively different from the corresponding van't Hoff enthalpies of solution, and a detailed explanation is provided.