This dataset contains raw and processed LC–MS/MS data generated for a prospective validation of label-free untargeted proteomics for host cell protein (HCP) quantification in biotherapeutic matrices. A stable isotope-labeled CHO whole-proteome standard was spiked into the NIST monoclonal antibody reference material (RM 8671) across seven concentration levels (20–80 ng total HCP per injection). Samples were analyzed using an Evosep One nanoLC system coupled to a Bruker timsTOF Pro operating in ddaPASEF mode. The study comprised four independent analytical assays totaling 198 injections and was designed to evaluate analytical performance characteristics under an ICH Q2(R2) total-error validation framework. The dataset supports evaluation of identification error control, quantitative trueness, precision, intermediate precision, and total error across the validated range. Raw Bruker timsTOF acquisition files and identification results generated with SpectroMine are provided together with the FASTA database and analysis parameters to ensure full reproducibility.