PurposeHypoxia contributes to radiotherapy resistance and more aggressive behaviour of several types of cancer. This study was designed to evaluate the repeatability of intratumour uptake of the hypoxia tracer [18F]EF5 in paired PET/CT scans.MethodsTen patients with newly diagnosed head and neck cancer (HNC) received three static PET/CT scans before chemoradiotherapy: two with [18F]EF5 a median of 7 days apart and one with [18F]FDG. Metabolically active primary tumour volumes were defined in [18F]FDG images and transferred to co-registered [18F]EF5 images for repeatability analysis. A tumour-to-muscle uptake ratio (TMR) of 1.5 at 3 h from injection of [18F]EF5 was used as a threshold representing hypoxic tissue.ResultsIn 10 paired [18F]EF5 PET/CT image sets, SUVmean, SUVmax, and TMR showed a good correlation with the intraclass correlation coefficients of 0.81, 0.85, and 0.87, respectively. The relative coefficients of repeatability for these parameters were 15%, 17%, and 10%, respectively. Fractional hypoxic volumes of the tumours in the repeated scans had a high correlation using the Spearman rank correlation test (r = 0.94). In a voxel-by-voxel TMR analysis between the repeated scans, the mean of Pearson correlation coefficients of individual patients was 0.65. The mean (± SD) difference of TMR in the pooled data set was 0.03 ± 0.20.ConclusionPretreatment [18F]EF5 PET/CT within one week shows high repeatability and is feasible for the guiding of hypoxia-targeted treatment interventions in HNC.