The inadequacy of single, simple arithmetic corrections for salicylate interference in the widely employed Glynn and Kendal technique for plasma paracetamol assay is highlighted by reference to an actual case of combined salicylate/paracetamol intoxication in an infant. Attention is drawn for the first time to the not insubstantial contribution to such interference made by even the minor metabolites of salicylate. The conclusion is reached that is necessary, particularly in the assessment of paracetamol toxicity, to determine the presence of salicylate, and when present, to employ a specific method for the estimation of paracetamol.