Clinical mastitis (CM) is the most prevalent and costly disease in dairy cows. Our objective was to develop for biomarker and pathway discovery a robust analytical workflow that can identify global serum metabolomic changes in healthy, close-up dairy cows that subsequently did (MastitisPost; n = 8) or did not (Control; n = 9) develop CM. Using a nested case-control design, we measured weekly serum signatures during the prepartal transition period and directly after calving. For analysis, we combined ethanol extraction, ultra-performance liquid chromatography coupled with quadrupole ion mobility time-of-flight mass spectrometry for data acquisition, and self-organizing maps for visualization of high-dimensional data. A repeated measures-in-time analysis of natural log-transformed data was conducted in PROC MIXED of SAS version 9.4. Fixed effects were collection time, group, and their interaction. Repeated measures within cows were modeled using a first-order heterogeneous variance-covariance matrix. Metabolomic analysis revealed global changes in amino acids and their metabolites, phospholipid precursors, acylcarnitines, and conjugated bile acids; their signal intensities were higher 21 d before calving and decreased stronger until calving in MastitisPost vs. Control cows. Complete separation of groups were observed at −21, −14, −7 d before and directly after calving between MastitisPost vs. Control cows for 18, 7, 0, and 1 of a total of 81 annotated metabolites, respectively. Free carnitine, trimethyllysine, proline, tyrosine, choline, phosphocholine and methylethanolamine phosphate had complete separation at the first 2 time points. The most consistent single biomarker was 3′ sialyl-lactose (signal intensities were 14.4-, 4.1-, 2.6-, and 2.3-fold higher at 21, 14, 7 d before and directly after calving, respectively, in MastitisPost vs. Control cows). We conclude that LC-MS metabolomic serum signatures indicate that global challenges in protein and lipid metabolism precede CM in transition dairy cows.