A multi-stage screening and identification scheme for the diagnosis of inborn errors in amino acid, fatty acid, carbohydrate and intermediary metabolism is described. The method is based on a computerized analysis of data obtained with dual-column gas chromatography/FID (GC/FID) and gas chromatography/mass spectrometry (GC/MS) profiling of urinary organic acids. It involves: (1) isolation of the compounds by solvent or solid-phase extraction; (2) conversion into trimethylsilyl (TMS) or TMS-oxime derivatives; (3) GC/FID analysis on SE-52 and OV-1701 capillary columns; (4) tentative identification by comparing the methylene unit (MU) values on both columns with a user-built library of reference compounds; (5) quantitative evaluation of the excretion profile; and (6) analysis by GC/MS of samples with an abnormal profile using an automated peak identification programme.