Abstract: Single-cell proteomics (SCP) aims to resolve protein heterogeneity among cells within tissue microenvironments. Single-cell proteomics analysis inherently relies on high-sensitivity mass spectrometry platforms. In this study, a series of optimizations for SCP analysis in a routine approach based on timsTOF Pro2 MS was reported. HeLa cells were selected as testing materials and processed to single cells through CellenONE. Single HeLa cells were treated under different experimental conditions, such as lysis buffer, cell disruption, and digestion period, to seek an optimized protocol with enhanced efficiency of protein extraction. For optimizing LC conditions, the samples with 5 ng were loaded onto the LC, and chromatographic parameters such as LC column type, elution gradient, flow rate, and column temperature were assessed. Mass spectrometer of timsTOF Pro2 was used under data-independent acquisition (DIA) mode, and the acquisition parameters were focused on m/z isolation window and ramp time. For protein extraction, an ammonium bicarbonate buffer system combined with n-Dodecyl-β-D-maltoside (DDM) and two-hour enzymatic digestion enabled an increase of the identified proteins. With the extraction system, analytical reproducibility reached a stable status, the overlap rate of inter-samples at 78%-82% and the correlation coefficients of protein quantification exceeding 0.88. In chromatography to separate 5 ng peptides, a 6 cm column was employed under a 15 min short gradient with a flow rate of 200 nL/min at temperature of 45 ℃. During the refinement of mass spectrometry parameters, the m/z isolation window was set at 125 ms ramp time. The optimized method achieved remarkable performance in protein identification at single-cell level, with an average of 2 459 proteins/HeLa cell. Through systematic optimization of the parameters of the timsTOF Pro2 platform, deep proteome identification can be achieved at the single-cell level. This study provides a set of optimizable solutions for laboratories equipped with similar high-performance LC-MS platforms, and preliminarily validates its applicability in multiple cell lines, offering methodological references for related basic research.