For patients with focal epilepsy scheduled for surgery, including MRI-negative cases, (18)FDG-PET was shown to disclose hypometabolism in the seizure onset zone. However, it is not clear whether grey matter hypometabolism is informative of the integrity of the surrounding white matter cerebral tissue. In order to study the relationship between metabolism of the seizure onset zone grey matter and the integrity of the surrounding white matter measured by diffusion tensor imaging (DTI), we performed a monocentric prospective study (from 2006 to 2009) in 15 children with pharmacoresistant focal epilepsy, suitable for interictal (18)FDG-PET, T1-, T2-, FLAIR sequence MRI and DTI. Children had either positive or negative MRI (eight with symptomatic and seven with cryptogenic epilepsies, respectively). Seven children subsequently underwent surgery. Standardised uptake values of grey matter PET metabolism were compared with DTI indices (fractional anisotropy [FA], apparent diffusion coefficient [ADC], parallel diffusion coefficient [PDC], and transverse diffusion coefficient [TDC]) in grey matter within the seizure onset zone and adjacent white matter, using regions of interest automatically drawn from individual sulcal and gyral parcellation. Hypometabolism correlated positively with white matter ADC, PDC, and TDC, and negatively with white matter FA. In the cryptogenic group of children, hypometabolism correlated positively with white matter ADC. Our results demonstrate a relationship between abnormalities of grey matter metabolism in the seizure onset zone and adjacent white matter structural alterations in childhood focal epilepsies, even in cryptogenic epilepsy. This relationship supports the hypothesis that microstructural alterations of the white matter are related to epileptic networks and has potential implications for the evaluation of children with MRI-negative epilepsy.