Abstract In expanded newborn screening programs by liquid chromatography/tandem mass spectrometry false negatives for tyrosinemia type I are a significant problem. We describe a method for inclusion of succinylacetone in order to avoid false negatives. We studied spots from 13 000 neonates born in Tuscany (January–May 2007) and ten spots from six patients with tyrosinemia type I. The traditional screening method was modified by adding dioxooctanoid acid (or 13 C 2 ‐succinylacetone) as an internal standard to the methanolic solution of deuterated acylcarnitines and amino acids. A hydrazine solution was added to the mixture. The times of extraction, butylation and drying were only slightly prolonged. Specific multiple reaction monitoring for derivatized and labelled succinylacetone and dioxooctanoic acid was carried out. The assays were linear up to 100 µmol/L for succinylacetone. Intra‐ and inter‐day imprecision data were in the range of 1.34% to 7.09% and 3.50% to 4.49%. Limits of detection and of quantification were 0.2 µmol/L and 0.4 µmol/L, respectively. Recovery ranged from 97.02% to 100.29%. Succinylacetone levels in samples from unaffected neonates were very close to the detection limit. Of the 46 recalls, eight (17.4%) were for abnormal tyrosine levels and all these cases had succinylacetone levels within the normal range (<2.4 µmol/L). In ten spots from six affected patients succinylacetone values ranged from 3.3 to 35.0 µmol/L. Including succinylacetone in newborn screening programs for amino acids and acylcarnitines avoids false‐negative results for tyrosinemia type I. Newborn screening laboratories should consider implementing these modifications. Copyright © 2008 John Wiley & Sons, Ltd.