Capturing phosphopeptides from complicated biological samples is essential for the discovery of new post-translational modification sites and disease diagnostics. Although several two-dimensional (2-D) materials have been used for phosphopeptides capturing, metal-organic framework (MOF) nanosheets have not been reported. The Ti-based MOF nanosheets have well-defined 2-D morphology, high density of active sites, large surface area, and an ultrathin structure. Phosphopeptides can be efficiently extracted and superior detection limits of 0.1 fmol μL-1 can be achieved even for an extremely low molar ratio of phosphoprotein/nonphosphoprotein (1:10000) mixtures. The selectivity over nonphosphopeptides can be enhanced further by pretreatment with a 10 mM salt solution (β-glycerophosphate disodium, NaCl, or KCl). The performance of 2-D Ti-based MOF nanosheets is much better than Zr-based MOF (Zr-BTB) nanosheets or any other Ti-based 3-D MOF counterpart, such as MIL-125 and NH2-MIL-125. The nanosheets were used for in situ isotope labeling for abnormally regulated phosphopeptides analysis from serum samples of type 2 diabetes patients. The relative quantitative results showed that three of the phosphorylated fibrinogen peptides A (FPA, DpSGEGDFLAEGGGV, DpSGEGDFLAEGGGVR, and ADpSGEGDFLAEGGGVR) were down-regulated, while the other isoform (ADpSGEGDFLAEGGGV) was up-regulated in the serum samples of type 2 diabetes patients compared with those of healthy volunteers. Finally, proteomics analysis showed selective enrichment of phosphopeptides with 2-D Ti-based MOF nanosheets from real samples, including tryptic digests of mouse brain neocortex lysate, mouse spinal cord lysate, and mouse testis lysate, followed by LC-MS/MS analysis. Total numbers of 2601, 3208, and 2866 phosphopeptides were successfully identified from the three samples, respectively. The 2-D Ti-based MOF nanosheets significantly improved sample preparation for mass spectrometric analysis in phosphopeptides and phosphoproteomics research.