To establish a metabolomic profile that could be used as a screening tool for possible deaths attributed to hypoglycaemia. Postmortem metabolomics can be used to assist death investigations by characterising new biomarker profiles differentiating causes of death. Hypoglycaemia-related deaths, particularly insulin intoxications, are difficult to identify without support from antemortem blood samples. The biological analysis of exogenous insulins in postmortem cases is troublesome, as immunoassays are not precise when handling haemolysed blood samples from postmortem cases. However, in recent years advances have been made in identification of insulin analogues, particularly in vitreous fluid. Nevertheless, due to such limitations it is probable that a proportion of hypoglycaemia-related deaths remain unknown. The aim of this investigation is to build a postmortem metabolomics model to be used as a tool to identify deaths related to hypoglycaemia. Data from ultra-high-performance liquid chromatography-quadrupole time-of-flight (UHPLC-QTOF) mass spectrometry was obtained from postmortem cases including; 19 insulin intoxication cases (hypoglycaemia group), 19 diabetic coma cases (hyperglycaemia group), and 38 hanging cases (control group). Additionally, a random test group of postmortem cases ( n = 726) were used for test screening for possible unidentified hypoglycaemic deaths. Mass spectrometry data was processed in XCMS using R. Orthogonal-partial least squares discriminant analysis (OPLS-DA), in SIMCA, was used to build models comparing metabolomic features between hypoglycaemia and control groups, and hypoglycaemia and hyperglycaemia groups. These two models were combined into a shared-unique-structures (SUS) plot, and the discriminating features specific for hypoglycaemia group were selected for identification and used to build the screening model. Features were identified using the HMDB and METLIN databases. XCMS data analyses resulted in the identification of 1775 metabolomic features. The OPLS-DA model for hypoglycaemia vs control resulted in complete group separation with R2 = 0.90 and Q2 = 0.62, and the model for hypoglycaemia vs hyperglycaemia resulted in complete group separation with R2 = 0.93 and Q2 = 0.86. The SUS plot of these two models resulted in 83 features that were considered to discriminate the hypoglycaemia group. An OPLS-DA model using only those 83 discriminating features was built, resulting in complete group separation of hypoglycaemia, hyperglycaemia, and control groups (R2 = 0.73, Q2 = 0.56). Group classification prediction was performed for the random test group on this model, which resulted in 46 (6.0%), 17 (2.2%), and 332 (43.5%) of cases being classified as belonging to hypoglycaemia, hyperglycaemia, and control groups, respectively. Cases classified as belonging to no group were 363 (47.6%), falling below the threshold of group classification in the model, and the remaining four cases were classified as belonging to two of the three groups in various combinations. The reported causes of death for those classified as hypoglycaemia showed a varied list of causes, including cardiovascular complications, substance overdose/poisoning, asphyxia, amongst others. However, six cases have an undetectable cause of death reported. The difficulties in the identification of hypoglycaemia-related deaths and the toxicological analysis of exogenous insulin remains. However, based on the results here we can see that postmortem metabolomics is able to discriminant deaths attributed to hypoglycaemia, hyperglycaemia, and non-glycaemic deaths (controls). Additionally, the model for postmortem case screening classified a discrete group of the random test group (only 6%) as hypoglycaemia, and whilst the individual cases still require review, this low number of classifications is a good indication that a viable screening model could be developed. Postmortem metabolomics is a useful tool in the screening of possible hypoglycaemia-related deaths. Further studies are required to expand on the potential of postmortem metabolomics as a useful tool in the investigation of hypoglycaemia-related deaths, and the future application of identifying potential insulin intoxications.