Wilms tumour is a great therapeutic success story within paediatric oncology; its prognosis is excellent. Although mainly sporadic, occurring in otherwise well children, it occurs in a small number of genetically predisposed children. Thus regular surveillance imaging is performed in predisposed children in parts of the USA and Europe. The risks and benefits of surveillance are unclear, as the existing ad-hoc surveillance protocols are lacking in consistency of practice and equity of provision. We present guidelines for Wilms tumour surveillance based on a review of current practice and available evidence, outlined by a multidisciplinary working group in the UK. Wilms tumours are bilateral in 4-13% of affected children. Bilateral synchronous nephroblastomas are observed in 5% of affected children and are usually associated with the presence of nephrogenic rests, congenital malformations and predisposing syndromes. The major challenge in bilateral disease is to achieve a cure and at the same time to preserve sufficient functional renal tissue for normal growth and development. The association among Wilms tumour, nephrogenic rests and nephroblastomatosis makes detection and characterization of renal lesions with imaging extremely important. We discuss the relative strengths and weaknesses of the different modalities used for diagnosis and follow-up in bilateral renal disease. We also discuss newly emerging diagnostic imaging tests such as (18)F-fluorodeoxyglucose positron emission tomography (FDG-PET). This technique, when fused with CT (PET-CT), allows accelerated metabolic activity to be accurately anatomically localised and so is potentially useful for staging, assessment of treatment response, and for surgical and radiotherapy planning. In addition, quantitative MRI techniques have been proved to be valuable in intracranial tumours, but no such role has been validated in abdominal disease. Diffusion-weighted imaging with calculation of ADC maps is feasible in abdominal tumours, and our own preliminary data suggest that tissue cellularity is an important determinant of ADC value, which might help in terms of early prediction of therapy response.