A trial involving 208 girls aged 5-12 who had bacteriuria on screening was carried out to determine the effects of covert bacteriuria on renal growth and scarring. 98 were observed for 4 years without treatment and 110 were treated. 77% of the treated girls, but only 26% of the untreated controls, were free of infection for at least half of the 4-year follow-up. However, treatment had no effect on the emergence of symptoms, clearance of vesico-ureteric reflux (V.U.R.), kidney growth, or the progression of kidney scars. New kidney scars did not develop in previously unscarred kidneys. 12 (6%) of the 208 girls had progression of re-existing kidney scars; this high-risk group was characterised by kidney scarring on the initial X-ray, a high prevalence of v.u.r., and persistent or recurrent bacteriuria. Screening for covert bacteriuria cannot therefore be recommended in schoolgirls since kidney damage associated with infection arises before the age of 5.