The changing pattern of antimicrobial resistance in the causative microorganisms of urinary tract infection (UTI) in childhood is a growing problem. The aims of this study were to assess the resistance patterns of urinary isolates to commonly used antimicrobials and to evaluate the options for empirical treatment of UTI. A prospective cross-sectional analysis of bacteria isolated from children with UTI was performed between January 2003 and January 2004. Resistance to antibiotics was analysed in three age groups: Group I, < or =12 months; Group II, 13-60 months; and Group III, >60 months. A total of 165 urinary pathogens were isolated from 131 patients. Mean patient age was 63.7+/-49.8 months. The most common causative agent was Escherichia coli (87% of cases) followed by Klebsiella pneumoniae (10%). Resistance to ampicillin (74.2%) and co-trimoxazole (61.3%) was significant in all isolates. Nitrofurantoin was the most active agent against E. coli (2.2% resistant isolates), followed by amikacin (4.9%), ceftriaxone (7.5%) and ciprofloxacin (12%). None of the isolates from Group I patients were resistant to ciprofloxacin and a low resistance rate (7.1%) was noted for amikacin. In Group II patients, none of the isolates were resistant to amikacin, and ceftriaxone was the second most suitable antibiotic (resistance rate 2.2%). In Group III patients, the lowest resistance rate was against nitrofurantoin (2.7%). In conclusion, we observed that the use of ampicillin and co-trimoxazole as a single agent for empirical treatment of a suspected UTI would not cover the majority of urinary pathogens in our region. Whilst amikacin, with a negligible resistance rate, was suitable in all age groups, gentamicin might still be useful as an empirical treatment of UTI in children aged >1 year. Nitrofurantoin could be included as a reasonable alternative in the empirical treatment of lower UTI in older children.