Tramadol (TR) metabolism is mainly dependent on the enzymatic activity of CYP2D6, which is controlled by genetic polymorphisms. Individuals are classified as poor (PMs), intermediate (IMs), extensive (EMs) or ultrarapid metabolizers (UMs) according to their genotype or phenotype. The determination of the metabolic phenotype for CYP2D6 can be of utmost importance in forensic and clinical contexts that involve TR intake. The present study aimed to describe CYP2D6 genetic variants in cases of TR-related deaths and to assess which metabolic ratio(s) (MRs) would allow to determine CYP2D6 phenotype without having to perform genetic analyses. Forty-eight postmortem blood samples were selected from TR-related death cases previously analyzed in a forensic context in North of France between 2013 and 2019. Initial available data included blood concentrations of TR and its two main metabolites (M1& M2) determined using a LC-MS/MS method. TR metabolism was expressed as various MRs comprising TR/M1, TR/M2 and M2/M1. After DNA extraction, sequencing was used for genetic variant detections that affect CYP2D6 activity/expression. In the present study, the allelic variants with the higher frequency were CYP2D6*1 (68%), followed by *4 (21%). The most frequent phenotype is EMs (59.6%), followed by IMs (23.4%), PMs (12.8%), and UMs (6.4%). There was no significant correlation between each calculated MR and the genotypically-predicted phenotypes, except for M2/M1 which appears related to the PM phenotype. The observed distribution of CYP2D6 genetic variants in this TR-related death population was similar to that found in the general Caucasian population. The present study displayed that the blood M2/M1 ratio could be the best-correlated tramadol MR to the PM phenotype, and could thus be used in forensic contexts where genetic analyses are not possible or poorly informative. For the other phenotypes, especially the UM phenotype, genetic analysis appears to be the only reliable method to predict the CYP2D6 phenotype.