The feasibility of developing a technique for rapid monitoring of potato chip quality by using infrared spectroscopy combined with multivariate analysis was investigated. Spectra of 15 commercial potato chips were collected by using an integrating sphere (near-infrared, NIR) or attenuated total reflectance (mid-infrared, ATR-MIR) accessories. Oil from potato chips was extracted with a carver press and spectra were collected by using a temperature controlled ATR accessory. The moisture and fat content were analyzed with reference methods AOAC #984.25 and AOAC #963.15, respectively. The spectral data were analyzed by partial least squares regression (PLSR) and calibration models generated. Fat content ranged from 18 to 45%, and moisture content ranged from 1.2 to 4%. Correlation coefficient (r) for moisture was >0.97 with standard error of cross validation (SECV) < 0.3% for both NIR and MIR techniques. Prediction models for fat had r > 0.96 and SECV < 1.60 for both NIR and MIR. Classification model based on Soft Independent Modeling of Class Analogy (SIMCA) differentiated potato chips by source of frying oil. MIR spectroscopy identified samples containing -trans fatty acids at 967 cm-1. A fast, simple and accurate method to evaluate potato chip quality was developed by using infrared spectroscopy. © 2008 Elsevier Inc. All rights reserved.