Cow milk contains essential nutrients for humans, and its bulk composition is usually analyzed using Fourier transform infrared spectroscopy. The higher sensitivity of nuclear magnetic resonance (NMR) spectroscopy can augment the extractible qualitative and quantitative information from milk to nearly 60 compounds, enabling us to monitor the health of cows and milk quality. Proton (¹H) NMR spectroscopy produces complex spectra that require expert knowledge for identifying and quantifying metabolites. Therefore, an efficient and reproducible methodology is required to transform complex milk ¹H NMR spectra into annotated and quantified milk metabolome data. In this study, standard operating procedures for screening the milk metabolome using ¹H NMR spectra are developed. A chemical shift library of 63 milk metabolites was established and implemented in the open-access Signature Mapping (SigMa) software. SigMa is a spectral analysis tool that transforms ¹H NMR spectra into a quantitative metabolite table. The applicability of the proposed methodology to whole milk, skim milk, and ultrafiltered milk is demonstrated, and the method is tested on ultrafiltered colostrum samples from dairy cows (n = 88) to evaluate whether metabolic changes in colostrum may reflect the metabolic status of cows.