Target deconvolution is essential for elucidating the molecular mechanisms, therapeutic efficacy, and off-target toxicity of small-molecule drugs. Thermal proteome profiling (TPP) is a robust and popular method for identifying drug-protein interactions. Nevertheless, classical implementation of TPP using isobaric labeling of peptides is tedious, time-consuming, and costly. This prompts the adoption of a label-free approach with data-independent acquisition (DIA), but with substantial compromise in protein coverage and precision. To address these shortcomings, we improvised a spike-in proteome strategy for DIA with TPP to counteract the reduction in protein quantity following sample heating. Protein coverage, data completeness, and quantification precision are significantly improved as result. Additionally, a calibration algorithm was developed to correct for spike-in effects on fold changes. The integration of DIA-TPP with the matrix-augmented pooling strategy (MAPS) to increase experiment throughput demonstrates performance comparable to that of existing TMT-TPP-MAPS. With this spike-in proteome strategy, we also successfully identified the thermal stabilization of CA13 by dorzolamide hydrochloride as well as GSTZ1 and tyrosyl-DNA phosphodiesterase 1 of opicapone that eluded detection without spike-in proteome.