Clinically-relevant proteins are routinely characterized by targeted proteomic methods, which offer high accuracy and reproducibility. However, assays developed for these techniques lack distinction between different alleles expressed in biological samples. The significance of assessing such variations in genes relevant to pharmacology will depend on their prevalence and effects on drug therapy. We propose quantitative allele-specific proteomics for simultaneous abundance measurement and determination of missense polymorphisms. We employed a targeted proteomic strategy using a QconCAT standard which included two surrogate peptides (at 1:1 ratio) for a prevalent variation of CYP2B6 (K262R) so that the two variants could be quantified directly. Measurement was carried out in 24 human liver samples, out of which 21 were genotyped. Allele-specific analysis of CYP2B6 expression was accurate and precise (CV < 9%), leading to determination of allele expression ratios (variant to wild type) for heterozygous (1.006 ± 0.079, n = 12) and homozygous (0.005 ± 0.004, n = 8) phenotypes. The abundance of CYP2B6 was 7.4 ± 7.8 pmol mg−1 microsomal protein and showed good correlation with activity against mephenytoin (Rs = 0.91, p < 0.0001; R2 = 0.93). Comparable abundance (and activity) appeared to be associated with genotypes that express at least one wild type allele, which was corroborated by turnover values. This proof-of-principle study demonstrates the possibility of simultaneous determination of CYP2B6 phenotype and abundance by independent assessment of allele products.