Cannabis sativa has a long history of domestication both for its bioactive compounds and its fibers. This has produced hundreds of varieties, usually characterized in the literature by chemotypes, with Δ9-THC and CBD content as the main markers. However, chemotyping could also be done based on minor compounds (phytocannabinoids and others). In this work, a workflow, which we propose to name cannabinomics, combines mass spectrometry of the whole metabolome and statistical analysis to help differentiate C. sativa varieties and deciphering their characteristic markers. By applying this cannabinomics approach to the data obtained from 20 varieties of C. sativa (classically classified as chemotype I, II, or III), we compared the results with those obtained by a targeted quantification of 11 phytocannabinoids. Cannabinomics can be considered as a complementary tool for phenotyping and genotyping, allowing the identification of minor compounds playing a key role as markers of differentiation.