The objective of this study is to evaluate by NIR imaging the homogeneity of process intermediates obtained with different process parameters during the development of a new pharmaceutical solid form. The process under investigation is a solid dosage form based on extrusion. The parameters are two kinds of crystallizations, two sizes of particle of API, two screw speeds during the extrusion and two milling screens used to reduce the extrudates into a granulate form. Two kinds of intermediates are evaluated: the extrudates and the cores. Two approaches are used to analyze the data: the univariate NIR analysis which consists in wavelength selection and multivariate analysis, i.e., Classical Least Squares (CLS), which takes into account the whole spectra. The univariate method reveals good chemical homogeneity of the extrudates but differences in their physical aspect. CLS shows well-distributed excipients for all the cores; differences in the sizes of the granules have also been revealed. The univariate method can be applied on simple chemical systems such as binary mixtures. When complex samples are analyzed, multivariate analysis is the method of choice. This study demonstrates that NIR imaging can be a useful tool for the optimization of the process and for the selection of the final parameters of the process.