The species-specific components (SSC) of animal signals can facilitate species recognition and reduce the risks of mismatching and interbreeding. Still, empirical evidence for SSCs in chemical signals is scarce and mostly limited to insect pheromones. Based on the proteinaceous femoral gland secretions of 36 lizard species (Lacertidae), we examine the SSC potential of proteins in lizard chemical signals. By quantitatively comparing the one-dimensional electrophoretic patterns of the protein fraction from femoral gland secretions, we first reveal that protein composition is species-specific, accounting for large part of the observed raw variation, and allowing us to discriminate species on this basis. Secondly, we find increased protein pattern divergence in sympatric, closely related species. Thirdly, lizard protein profiles show a low phylogenetic signal, a recent and steep increase in relative disparity, and a high rate of evolutionary change compared to non-specifically-signal traits (i.e. body size and shape). Together, these findings provide support for the species-specificity of proteins in the chemical signals of a vertebrate lineage. We made available two datasets: rawdata.Rdata is an R-data file containing a single R-list object called “epg” (which stands for “electrophoretograms”). The list has 7 entries: $profile The normalised luma values for each sample individual; row names correspond to the species name (6 letter abbreviations) followed by a progressive integer $hpx a progressive number indicating the molecular weight integer corresponding to each luma value in epg$profile $kda the corresponding average molecular weight for each bin (in kDa) $spe a factor indicating the species for each profile row $gel a factor indicating the gel from which the profile was extracted $abso the absorbance value, proxy for protein concentration for each sample $epg.species the geometric mean of the EPG for each species (row names) gel.zip is a zipped file with all gel images used to extract individual EPG. All other information and data needed to replicate our analysis are available in the published paper or in reference therein.