SummaryThe aim of this work was to develop a fast, versatile, inexpensive and environmentally safe analytical method to quantify simple sugars, malic acid and total phenolic compounds in apple pomace, considering its potential use as a raw material with value instead of as an industrial waste. Diffuse reflectance infrared spectroscopy (DRIFTS) measurements of twenty-six samples of apple pomace were analysed by partial least squares regression (PLSR), using several signal pre-processing methods. Multivariate models developed with four to five latent variables (LVs) and based in the MIR (mid-infrared) region had good prediction for the determination of sucrose, fructose, malic acid and total phenolic compounds, with average errors between 3.9% and 6.6%. By contrast, glucose was better determined by models developed in the NIR (near-infrared) region and using six LVs, yielding an average error lower than 7.4%. These results confirmed the feasibility of the multivariate spectroscopic approach as an alternative for expensive and time-consuming conventional chemical methods.