Postmortem drug diffusion from gastric residue was assessed in a human cadaver model. Fifty milligrams of amitriptyline (Ami) and 5 g of paracetamol (Par) suspended in 350 ml of 10% methanol, 0.1 N HCl, and 50 ml urograffin with 5 g lithium carbonate (alkaline model) or without lithium (acidic model) was instilled into the stomach through an esophageal tube via a neck dissection. Multiple samples were obtained after 48 h at room temperature (range in mean hourly room temperature: 15.6-20.7°C, n = 9). The pH of the gastric contents (alkaline model range = 8.3-8.9, n = 5; acidic model range = 3.4-3.8, n = 5) had no significant effect. Drug diffusion was most marked in the left lung base, with drug concentrations (μg/g) of 0.1-13.9 for Ami, 65-524 for Par, and 13-161 for lithium. Similarly affected were the left lobe of the liver (Ami, 0.1-54.9; Par, 7-218; lithium, 7-39), the spleen (Ami, 0.6-24.3; Par, 104-663; lithium, 27-106), and pericardial fluid (Ami, 0-4.5; Par, 48-641; lithium, 12-56). Diffusion into gallbladder bile, cardiac blood, aortic blood, and blood of the inferior vena cava was less severe. The left kidney and left lung were more severely affected than the right kidney and lung, and similarly the left and right psoas muscles. Least affected was the right anterior lobe of the liver and the lung apexes. This phenomenon may significantly influence drug concentrations in liver and in blood samples obtained from the torso, and consequently liver/blood drug ratios. To circumvent the problem of postmortem drug diffusion from the stomach, it is recommended that blood be sampled from a peripheral vessel, skeletal muscle from a limb, liver from deep within the right lobe, and lung from the apex rather than the base.