Fifty waste cooking oils (WCOs) were transesterified with methanol (1: 8 WCO: methanol molar ratio) at 60 degrees C for 60 min using NaOH as catalyst (1% wt.). Fatty acid methyl ester (FAME) composition of the resulting biodiesels was analysed by gas chromatography, and near infrared (NIR) spectra of these biodiesels and those of the starting WCOs were acquired. Biodiesel cetane number was then calculated from both FAME composition and from biodiesel NIR spectra, this last technique using the former one as reference data. Because of transesterification does not modify fatty acid distribution of the starting WCO, and the similarity between biodiesel and WCO NIR spectra, biodiesel cetane number was successfully predicted from WCO NIR spectra, achieving RPD (ratio of performance to deviation) of 3.83. Therefore, biodiesel cetane number (and, as consequence, any other biodiesel property related to FAME composition) can be predicted by NIR spectroscopy before performing the transesterification reaction, which allows beforehand selecting the most suitable substrates for biodiesel production.