This prospective study from India included 744 patients of whom 112 had renal function less than or equal to 20% at the time of diagnosis. Thirty percent underwent a nephrostomy initially. Ten with no function had a nephrectomy. Of the 102 with remaining kidneys, differential function was 0-9% in 40, 10-20% in 62. Follow-up of one to eight years (mean 4.6 ± 1.34 years) showed significant improvement in drainage in all 96 patients that were available for follow-up. Differential function improved and remained stable in these cases between the initial postoperative three-month renogram and scans up to eight years later. The authors conclude that, whereas previously kidneys with function less than 15-20% were removed rather than having a pyeloplasty performed, their data suggests that nephrectomy should only be considered in children with 0-5% function or, if there is no renal parenchyma of note, if there is pyonephrosis, hypertension or minimal output on nephrostomy drainage. This policy will undoubtedly ‘save’ more kidneys.