Updated project metadata. The study describes comprehensive proteomics data of the cholesteatoma disease. Cholesteatoma is a serious and destructive growth of keratinizing squamous epithelium in the middle ear, with a poorly understood etiopathogenesis. The large scale proteomics approach used in this study highly expands our molecular knowledge of cholesteatoma and implicates several biological functions in its pathology. Human samples of cholesteatoma, neck of cholesteatoma, tympanic membrane, external auditory canal skin, and middle ear mucosa were analyzed. Approximately two thousand unique proteins were identified by label-free nanoLC-MS/MS proteomics. The protein and peptide identification data were obtained by Proteome Discoverer (Thermo Scientific, version 1.3.0.339) processing of LTQ Orbitrap raw files using the Mascot algorithm (Matrix Sciences, version 2.4.0). To yield comprehensive proteome data the tissue biopsy was separated into ten protein fractions, which were MS analysed separately, and the resulting MS files were subsequently merged in the database analysis. Precursor mass tolerance was 5 ppm and fragment mass tolerance was 0.5 Da. Dynamic Modification was Oxidation (M) and Static Modification was Carbamidomethyl (C); annotated in the PRIDE files as mass deviations (e g for double charged peptides m/z deviation of approximately 8.00 and 28.51 for modification of Met and Cys, respectively). FASTA file was SwissProt_2012_03.fasta (containing 20,255 Homo sapiens sequences). MudPit scoring was applied, and up to two missed trypsin cleavages were accepted. Peptide cut off Score was 10 and protein relevance threshold was 20.