26 children, with antenatally detected, isolated, bilateral hydronephr
osis (thus consistent with pelvi-ureteric junction obstruction), were
investigated for up to 7 years. The aim was to avoid operation. 34 kid
neys were never operated on. Despite initially often severe hydronephr
osis and poor isotope wash-out on the renogram, half of the kidneys im
proved or even normalized during follow-up. The final glomerular filtr
ation rate (GFR) was always normal or almost normal. 18 kidneys came t
o surgery, because of poor filtration, symptoms or social reasons. At
the end of the follow-up, GFR was normal in all but 2 kidneys. In a su
bgroup of the kidneys, with initially severe reduction in filtration,
half were operated on and half were not; at the end, the GFR was norma
l in all of the non-operated kidneys (and in all but 1 of the operated
kidneys). Total GFR was at the end normal or almost normal also in th
e children with persistingly filtration-reduced kidneys (always unilat
erally), because of compensation by the contralateral, although hydron
ephrotic kidney. Thus, in this study the renographic parameters were n
ot very helpful in predicting the prognosis. There was no consistent d
ifference between operative and nonoperative treatments on final filtr
ation. Also hydronephrotic kidneys possessed a substantial capacity fo
r contralateral compensation. It is concluded that the initial approac
h to bilateral, antenatally diagnosed hydronephrosis may safely be non
-operative. This allows time for the clear expression of sustained imp
airment of renal filtration requiring surgical intervention.