Total blood carbon monoxide (TBCO) showed promising results in improving accuracy of CO determinations in blood and presenting better stability to different storage conditions. Therefore, it was proposed as an alternative biomarker to carboxyhemoglobin (COHb) for CO poisoning diagnosis. However, given that current interpretation reference values exist for COHb only, it is difficult to implement TBCO analysis in routine. Therefore, we aimed at determining TBCO reference values for postmortem CO poisoning cases. A previously validated method for TBCO analysis via gas chromatography-mass spectrometry was applied to cardiac, peripheral, cranial and spleen blood samples collected from 92 autopsies. Autopsy cases included 21 non-CO-related and 71 CO-related cases with varying postmortem intervals (PMIs). Statistical analyses were performed using statistical software R Studio. When comparing lower to higher PMIs for non-CO-related cases, no significant differences were found, which suggests that CO formation or degradation at low PMIs does not occur. Spleen blood showed potential as an alternative matrix to CO determinations in cases with sample availability issues but needs to be evaluated for CO-positive cases. Results for cardiac blood in CO-related autopsies showed a positive correlation between COHb and TBCO values (R = 0.78). This value is lower than what is found in the literature, suggesting that even though COHb and TBCO are correlated, a potential underestimation of the true CO exposure might occur if only COHb values are taken into consideration. Samples were divided into CO exposure groups based on COHb concentrations, and with the data obtained, classification into the following TBCO concentration groups is proposed: no significant CO exposure case <6 µmol/mL, medium CO exposure case 6-20 µmol/mL and high CO exposure case >20 µmol/mL. Even if a higher number of samples in each group would enable to increase the confidence, these results are very promising and highlight the importance of TBCO measurement.