erroneous citing Piotr Kusnierczyk, Ludwik Hirszfeld Institute of Immunology and Experimental Therapy, Polish Academy of Sciences, Wroclaw, Poland 22 March 2007 Dear Editor, In a recently published article in the BMC Medical Genetics (2007, 8:3) Sofia Mayans and colleagues definitely prove a lack of association between CTLA-4 gene CT60 single nucleotide polymorphism (SNP) and serum level of soluble CTLA-4 (sCTLA-4). This is an interesting and valuable paper. However, in the Discussion, the Authors say that their results "are supported by data from Anjos et al.(2002) who also found no effect of the CT60 polymorphism on the expression of sCTLA-4". In the matter of fact, Anjos et al. did not study CT60 at all, but only +49A>G SNP. In contrast, the lack of effect of CT60 on sCTLA-4 serum levels was described earlier by Purohit et al (2005) who is cited by Mayans et al., but only for the lack of association of the sCTLA-4 level with type 1 diabetes. Competing interests I have no competing interests.