In a longitudinal prospective study 58 schoolgirls with covert bacteriuria were followed up for an average of 11.2 years (range 8.8 to 13.5 years). Intravenous urography was carried out at the start of the study (aged 4 to 11 years) and after completion of the follow up period (aged 14.8 to 22.3 years). After random allocation 27 of these girls received intermittent treatment for covert bacteriuria for the first four years and the control group received no treatment. The effect of covert bacteriuria, treatment, vesicoureteric reflux, and reflux nephropathy at presentation on the final renal length, progression of scarring, and development of new scars was analysed. No new scars were found in girls with bilaterally normal kidneys. In girls with reflux nephropathy, three kidneys showed progression of existing scars and two kidneys developed new scars. It was shown that final renal length was not influenced by vesicoureteric reflux or treatment, but reduced renal length at final assessment was associated with the presence of kidney scarring at initial assessment.