Second-generation biodiesel manufactured from waste cooking oils (WCO) and inedible animal fats (AF) are one of the alternatives to the first generation (1G) vegetable oil-based biodiesel. In this study, a quality control method is proposed to evaluate methanol content in waste fat methyl esters and is based on near infrared spectroscopy (NIR) combined with multivariate analysis. More specifically, calibration models are constructed using partial least squares regression (PLS) for the prediction of methanol content in rapeseed oil methyl ester (ROME), waste cooking oil methyl ester (WCOME), chicken fat methyl ester (CFME) and pork fat methyl ester (PFME) by Vis-NIR spectrometer. The calibration models are based on the absorbance spectra and computed data from five wavelength regions of 400-2170 nm, 780-2170 nm, 1400-2170 nm, 1400-1600 nm and 1970-2170 nm. For the cases with the highest prediction ability obtained in this study, the coefficient of determination of the model's goodness-of-fit for methanol concentrations range 0-5% (v/v) was R2 > 0.990, and for concentrations 0-1% (v/v) was R2 > 0.994, indicating the spectroscopic approach effectiveness in methanol content detection relevant to the biofuel quality assessment. A pseudo-univariate limits of detection (LODpu) and quantification (LOQpu) as well as ratio of performance to deviation (RPD) were used to confirm the validity and to evaluate the practical applicability of developed models. In addition, the obtained results indicate the possibility of developing a transmission sensor for online monitoring of the production process and the quality of biofuel.