Milk powder is a food widely used by society and target fraudulent processes. In this way, it is of outmost importance the development of analytical methodologies for identification and quantification of such frauds. In this work will be presented fraud by adulteration in milk powder by potential chemical adulterants including urea, whey and sucrose. For this goal, this work uses Raman spectroscopic imaging in conjunction to multivariate curve resolution with alternating least squares method (MCR-ALS) for data treatment. The spectra were obtained using Raman Station 400F, from Perkin Elmer, laser of 785 nm, pixel of 50 um, range of 3200 -200 cm-1 with resolution of 2 cm-1. The area scanned was 2,4mm2, using 25 x 25 pixels and obtained 625 pixels in total. Data were arranged in augmented matrix, where several images are processed together, and individually, where each image is processed at a time, for further processing. It was studied the constraints: non-negativity, both in concentration in the spectral profile, and correlation. From this work, it was possible to identify adulterants in unknown composition samples, using SIMPLISMA (simple- to-use interactive self -modeling mixture analysis) to generate initial estimates of spectral profiles. Also, the quantification was performed, using only 6 standard samples, even in the presence of non-calibrated interferences, reaching the second order advantage. The predicted values for concentrations showed absolute error generally less than 5%, even for the most complex mixtures with three adulterants in both data processed individually as in augmented matrix. The repeatability of the data was verified by performing analysis in triplicate samples adulterated with sucrose, with standard deviations less than 3%. The models built from Raman hyperspectral imaging with MCR-ALS was able to detect and quantify the adulterants used in all samples studied.