FT-Raman spectroscopy has been used to investigate interactions between lecithin and cholesterol. Raman spectra of lecithin show multiple peaks which can be classified into three regions: hydrophobic chain, interfacial, and headgroup regions. Binary lipid mixtures (1:1, w/w, lecithin:cholesterol) were prepared by physical mixing, granulation, coprecipitation, hydration and heating (65 °C), and heating (120 °C). Regardless of the preparation method, no changes in the spectra were observed in the hydrophobic region. A shift in the wavenumber of the choline methyl asymmetric stretching mode was observed when the samples were prepared by coprecipitation, hydration and heating (65 °C), and heating (120 °C). This may indicate a modification of phospholipids in the headgroup region in these samples. The difference in degrees of frequency shift (physical mixing ≈ granulation < coprecipitation ≈ hydration and heating (65 °C) < heating (120 °C)) suggests that different levels of hydrogen bonding may have occurred in mixtures prepared with these methods. Multivariate analysis utilizing partial least squares regression based on selected wavenumber ranges was applied for the quantitative analysis of the amount of lecithin in lipid mixtures. Calibration models from physical mixing and heating (120 °C) exhibited lower R2 and root mean square error of cross validation (RMSECV) values compared to the other models suggesting lower sample homogeneity for these preparation methods. Low values of the mean absolute residues and mean Mahalanobis distances imply that the calibration model generated from physical mixing samples may be appropriate for quantitative analysis of lecithin in lipid mixtures prepared by any of the other techniques.