Antimicrobial prophylaxis using cefaclor or trimethoprim-sulfamethoxazole (co-trimoxazole) is recommended for children with vesicoureteral reflex (VUR) to prevent recurrent urinary tract infection (UTI). This retrospective study was performed by reviewing the data of children >5 years of age treated for recurrent UTI in six hospitals from 2010 to 2015. The criteria for UTI diagnosis is fever (>38°C) and positive results in urine culture (>104 colony-forming units/ml in midstream or withdrawn urine specimens). In total, 41 children were reviewed, and 31 children had recurrent UTI without antimicrobial prophylaxis and 10 had breakthrough (BT)-UTI treated with prophylaxis using cefaclor or co-trimoxazole. In the cases of BT-UTI treated with prophylaxis, 5 children received cefaclor and 5 received co-trimoxazole. We collected data on pathogens, antimicrobial resistance, and antimicrobial agents chosen for the empirical treatment of recurrent UTI. We also evaluated the validity of empirical therapy for recurrent UTI in this study. Various pathogens were found in children who received prophylaxis with cefaclor. The rate of empirical antimicrobial agents that were inappropriate based on antimicrobial susceptibility tests was higher in children who received prophylaxis with cefaclor (60.0%) than in those who received no prophylaxis (25.9%) or prophylaxis with co-trimoxazole (20.0%). Prophylaxis with cefaclor was found to be a risk factor for inappropriate empirical treatment in BT-UTI cases. The results suggest that the choice of empirical antimicrobial agents in BT-UTI cases should be carefully considered before treatment with prophylaxis. To encourage the adequate use of antimicrobial agents, we recommend prophylaxis with co-trimoxazole to prevent recurrent UTI.