SummaryTo provide further evidence for the lipid sink theory, we have developed an in vitro model to assess the effect of Intralipid Ò 20%on methaemoglobin formation by drugs of varying lipid solubility. Progressively increasing Intralipid concentrations from 4 to 24 mg.ml )1 suppressed methaemoglobin formation by the lipid soluble drug glyceryl trinitrate in a dose-dependent manner (p < 0.001). Both dose and timing of administration of Intralipid to blood previously incubated with glyceryl trinitrate for 10 and 40 min resulted in significant suppression of methaemoglobin formation (p < 0.0001 and p < 0.05, respectively). Mathematical modelling demonstrated that the entire process of methaemoglobin formation by glyceryl trinitrate was slowed down in the presence of Intralipid. Intralipid did not significantly suppress methaemoglobin formation induced by 2-amino-5-hydroxytoluene (partially lipid soluble) or sodium nitrite (lipid insoluble; both p > 0.5). This work may assist determination of the suitability of drugs taken in overdose for which Intralipid might be deployed.