Despite the increasing popularity of data-independent acquisition workflows, data-dependent acquisition (DDA) is still the prevalent method of LC-MS-based proteomics. DDA is the basis of isobaric mass tagging technique, a powerful MS2 quantification strategy that allows coanalysis of up to 10 proteomics samples. A well-documented limitation of DDA, however, is precursor coselection, whereby a target peptide is coisolated with other ions for fragmentation. Here, we investigated if additional peptide purification by traveling wave ion mobility separation (TWIMS) can reduce precursor contamination using a mixture of Saccharomyces cerevisiae and HeLa proteomes. In accordance with previous reports on FAIMS-Orbitrap instruments, we find that TWIMS provides a remarkable improvement (on average 2.85 times) in the signal-to-noise ratio for sequence ions. We also report that TWIMS reduces reporter ions contamination by around one-third (to 14-15% contamination) and even further (to 6-9%) when combined with a narrowed quadrupole isolation window. We discuss challenges associated with applying TWIMS purification to isobaric mass tagging experiments, including correlation between ion m/z and drift time, which means that coselected peptides are expected to have similar mobility. We also demonstrate that labeling results in peptides having more uniform m/z and drift time distributions than observed for unlabeled peptides. Data are available via ProteomeXchange with identifier PXD001047.