Abstract Objectives To compare the ultrasound characteristics with clinical features, final diagnosis and outcome; and to evaluate the halo size following glucocorticoid treatment in patients with newly diagnosed GCA. Methods Patients with suspected GCA, recruited from an international cohort, had an ultrasound of temporal (TA) and axillary (AX) arteries performed within 7 days of commencing glucocorticoids. We compared differences in clinical features at disease presentation, after 2 weeks and after 6 months, according to the presence or absence of halo sign. We undertook a cross-sectional analysis of the differences in halo thickness using Pearson’s correlation coefficient (r) and Analysis of Variance (ANOVA). Results A total of 345 patients with 6 months follow-up data were included; 226 (65.5%) had a diagnosis of GCA. Jaw claudication and visual symptoms were more frequent in patients with halo sign (P =0.018 and P =0.003, respectively). Physical examination abnormalities were significantly associated with the presence of ipsilateral halo (P <0.05). Stenosis or occlusion on ultrasound failed to contribute to the diagnosis of GCA. During 7 days of glucocorticoid treatment, there was a consistent reduction in halo size in the TA (maximum halo size per patient: r=−0.30, P =0.001; and all halos r=−0.23, P <0.001), but not in the AX (P >0.05). However, the presence of halo at baseline failed to predict future ischaemic events occurring during follow-up. Conclusion In newly diagnosed GCA, TA halo is associated with the presence of ischaemic features and its size decreases following glucocorticoid treatment, supporting its early use as a marker of disease activity, in addition to its diagnostic role.