The case of a patient with neonatal propionic acidaemia is reported. Despite an initially favourable response to the administration of an artificial formula, the patient finally died when 9.5 months old. Protein tolerance never exceeded 1.3 g kg-1 day-1. During the remission periods, when ingesting the formula, 3-hydroxypropionic acid was excreted alone or together with tiglylglycine and/or methylcitrate. In a period of ketoacidosis, in addition to these three metabolites and those of ketoacidosis, elevations of 2-methyl-3-oxovaleric acid, propionylglycine and 2-methyl-3-hydroxybutyric acid were found. A severe deficiency of propionyl-CoA carboxylase in cultured fibroblasts was detected; biotin, when added to the fibroblasts culture media, did not stimulate this enzyme activity. The effectiveness of the administered formula is discussed.