Acid value (AV) is an important parameter to illustrate the quality as well as degree of refining of peanut oil. A rapid near-infrared reflectance spectroscopy (NIRS) method was applied to determine AV in peanut oils. A partial least squares (PLS) regression model with a coefficient of determination (R(2)) of 0.9725 and a square error of cross-validation (SECV) of 0.308 was obtained. The prediction set gave a coefficient of determination (r(2)) and standard error of prediction (SEP) of 0.9379 and 0.333. Regarding qualitative evaluation, the classification of qualified peanut oil (with an acid value of less than or equal to 3 mg/g) and unqualified peanut oils (with an acid value of more than 3 mg/g) was conducted by using discriminant partial least squares analysis (DPLS). The results showed that DPLS technique was an effective method of classification model building, with a high correct percent of 96.55%. (C) 2009 Elsevier Ltd. All rights reserved.