Laura M.Labay* | AyakoChan-Hosokawa* | Joseph W.Homan* | Matthew M.McMullin* | Francis X.Diamond* | Mark M.Annand* | Stephanie M.Marco* | Jonn M.Hollenbach†


  1. * NMS Labs 200, Welsh Road Horsham, PA 19044
  2. † Berks County Medical Examiner and Coroner, 1047 MacArthur Road, Reading, PA


Poison hemlock (Conium maculatum L.) is a weed that grows rampant in many areas of North America. Forensic toxicology laboratories rarely receive requests to analyze biological specimens for the presence of poison hemlock. This report discusses two postmortem cases that were encountered over a decade apart and describes different analytical approaches that may be used to quantify coniine, a primary poison hemlock alkaloid, in biological specimens. The first case is from 2004 and involves a 27-year-old female that was found deceased in a relatively isolated area of California. Based on the presence of plant material at the scene and signs of its ingestion at autopsy, the possibility of hemlock poisoning was considered. Toxicological testing of the blood and gastric content by quantitative selected-ion monitoring Gas Chromatography/Mass Spectrometry (SIM-GC/MS) revealed the presence of coniine at concentrations of 410 ng/mL and 9300 ng/mL, respectively. The second case is from Pennsylvania and was sent for analysis in the spring of 2019. In this case, a male in his forties was found deceased in the kitchen area of a camper. Green substances, in liquid and residue forms, were observed in the sink. Mixtures of leaf-like material were also found in several bowls and pans. Subclavian blood screened positive for coniine by full-scan Gas Chromatography/Mass Spectrometry (GC/MS). Semi-quantitative confirmation testing was performed by Liquid Chromatography Tandem Mass Spectrometry (LC-MS/MS) and showed the presence of coniine at a concentration of 35 ng/mL. These analytical approaches can be used to substantiate or exclude poison hemlock exposure as a cause of death.

If you are interested in the full article, it is available at: Article's Abstract at