Sir, We read the paper by Levy et al1 with interest, however, we believe that it contains several weaknesses which should be discussed. As it is well known the diagnosis of GCA is based on the combination of clinical symptoms, laboratory test (CRP and ESR), and histopathology.2 The treatment must be started immediately, thus the temporal artery biopsy result usually confirms presumed diagnosis made on the basis of clinical picture and laboratory tests. It is well known that laboratory tests in GCA have some limited sensitivity: CRP 86.9% and ESR 84.1%.2 Both combined gave specificity 97%, but it was shown and discussed in several studies that they might be normal in GCA:2, 3, 4, 5 ESR was normal in 5–30% of patients with GCA.3 The authors1 inappropriately stated that ‘including their case there are only three published cases of isolated CRP-negative GCA and only two cases of simultaneous ESR and CRP negativity' as they did not notice nor analyze many similar cases (Table 1). Noteworthy, it was postulated that elevated ESR with normal CRP might occur even in 1.7% of GCA patients,3 and that these markers can be normal at the early stage of the disease.4 Table 1 Some publications reporting normal ESR and/or CRP Having in mind problems with ESR and CRP in GCA, the role of clinical picture of the disease seems to be very important. They include usually headache, tenderness of the scalp, jaw claudication and some systemic symptoms, including malaise, fever, anorexia, and weight loss.5 The symptoms are based on ischemia and/or inflammation. Thus, it is hard to understand why in the discussed case none of the clinicial symptomes was presented. Concluding, we agree that diagnosis of GCA might be problematic, but it must be based on clinical picture and laboratory tests, including ESR and CRP, and confirmed by temporal artery biopsy. GCA with normal ESR and/or normal CRP level is rare, but many cases of this form of disease were already described.