Insulin determination in blood sampled during post-mortem investigation has been repeatedly asserted as being of little diagnostic value due to the rapid occurrence of decompositional changes and blood haemolysis. In this study, we assessed the feasibility of insulin determination in post-mortem serum, vitreous humour, bile, and cerebrospinal and pericardial fluids in one case of fatal insulin self-administration and a series of 40 control cases (diabetics and non-diabetics) using a chemiluminescence enzyme immunoassay. In the case of suicide by insulin self-administration, insulin concentrations in pericardial fluid and bile were higher than blood clinical reference values, though lower than post-mortem serum concentration. Insulin concentrations in vitreous (11.50 mU/L) and cerebrospinal fluid (17.30 mU/L) were lower than blood clinical reference values. Vitreous insulin concentrations in non-diabetic control cases were lower than the estimated detection limit of the method. These preliminary results tend to confirm the usefulness of insulin determination in vitreous humour in situations of suspected fatal insulin administration. Additional findings pertaining to insulin determination in bile, pericardial, and cerebrospinal fluid would suggest that analysis performed in post-mortem serum and injection sites could be complemented, in individual cases, by investigations carried out in alternative biological fluids. Lastly, these results would indicate that analysis with chemiluminescence enzyme immunoassay may provide suitable data, similar to analysis with liquid chromatography-tandem mass spectrometry (LC-MS/MS) and immunoradiometric assay, to support the hypothesis of insulin overdose. Copyright © 2015 John Wiley & Sons, Ltd.