Diagnosing of hazelnut allergy was recently enhanced by value of serum specific IgE versus recombinant fractions, Cor a 1 and Cor a 8. Their positivity could be linked with clinical benignity for Cor a 1 and severity for Cor a 8. The authors report four cases contradicting this aphorism. The oral food challenge to hazelnut induced in these patients a severe clinical response with a severe increase of serum tryptase at the end of the challenge. The level of serum specific IgE to Cor a 8 was negative (or very low). For Cor a 1 the level was very high or low. We discuss the clinical relevance of the aphorism "Cor a 1 = benignity; Cor a 8 = severity", which use should be modulated in clinical practice. Other allergenic fractions could account for these cases. (c) 2012 Elsevier Masson SAS. All rights reserved.