The prevention of reflux nephropathy by screening will likely prove to be an elusive aim. Possible non-invasive screening tests include urinary analysis for microproteins, colour Doppler ultrasound and real-time ultrasound. However, it is likely that real-time ultrasound, in practice the only feasible screening tool, would fail to detect the large group of children whose vesicoureteric reflux (VUR) does not give rise to detectable dilatation. Routine ultrasound scanning of the fetal urinary tract in pregnancy constitutes a de facto screening programme. However, it has become apparent that prenatal ultrasound results in the detection of high-grade VUR in a population of boys rather than in the numerically more important population at risk: girls with low-grade VUR. Investigating infants with mild pelvic and pelvicaliceal dilatation for VUR might improve the sensitivity of prenatal ultrasound as a screening test, but the potential benefits of detecting VUR in this group of infants would have to be weighed against unnecessary micturating cystography in a large number of normal children. Presently, considerable clinical and experimental evidence indicates that urinary infection plays the major role in the aetiology of reflux nephropathy. By detecting asymptomatic VUR in early childhood, an effective screening programme could be expected to reduce the considerable burden of morbidity that is associated with the condition. Unfortunately, a reliable screening test remains a distant goal.