We read with great interest the article written by Karadaf and colleagues [1] entitled "Severity of carpal tunnel syndrome assessed with high frequency ultrasonography".The authors investigated the relationship between neurophysiological and ultrasound (US) evaluation in a sample of patients aVected by carpal tunnel syndrome (CTS) and they found agreement between the two measures: the increase of median nerve cross-sectional area (CSA) at the wrist reXects the damage severity as assessed by nerve conduction (NC) study.We agree with the results found by the authors and according to the trend showed in the literature, there is an increased interest on this topic [2-4] that demonstrates that US is able to detect median nerve damage in a very easy, non-invasive and non-expensive way.Another important value of US is that it provides not only a diagnosis of CTS, with a determination of CTS severity, but also able to supply morphological information about the nerve surrounding thus giving indirect data about the therapeutical approach.What is surprising in the article is that in the half patients aVected by clinical CTS, the neurophysiological examination was negative (49/99 symptomatic wrist) and these data are also followed by US evaluation. We have some comments on it.As American Academy of Neurology (ANN) previously [5] stated the diagnosis of CTS is mostly clinical with the morphofunctional test aimed in conWrming the diagnosis. Now, it is well accepted that, in order to increase the sensitivity of conventional conduction studies (sensory digit-wrist and motor wrist-thenar), segmentary and/or comparative tests should be used as stated in AAN and American Association of Electrodiagnostic Medicine recommendations [6]. When the standard tests yield normal results ("standard negative" hands), studies as follows could increase the electrodiagnostic sensitivity: 1. Segmental motor or sensory conduction tests in palmwrist segment. 2. Comparative studies (median-ulnar or median radial) 3. Segmental/comparative studies (as disto-proximal ratio)A study conducted on CTS patients showed that the sensitivity of standard tests can reach 83.5%, comparative/segmental tests can disclose abnormal Wndings in a further 11.4% of cases, providing CTS electrodiagnosis in about 7 of 10 "standard negative" cases. The overall sensitivity of protocol reaches in that way 94.9% [7]. The concordance of negativity of US with NC probably is due to the same limitation. Since the Wrst applications of US in the peripheral nerve study, reWnement has been done to improve the sensitivity and speciWcity of the evaluation. For example, the wrist-to-forearm ratio of median nerve area can be considered more sensitive than measure of median nerve area at the wrist alone, and should be used in case of mild impairment [8]. Moreover, the comparison of CSA measurements of the median nerve obtained at the level of the carpal tunnel with those obtained more proximally, at the level of the pronator quadratus muscle showed an increased accuracy of the diagnosis [9]...