BACKGROUND: Choosing an empiric treatment for urinary tract infections (UTIs) requires epidemiologic data on antibiotic nonsusceptibility (ANS) rates, and risk factors for UTI and ANS in the individual patient. We assessed ANS in community-acquired UTI, and its association with risk factors exposure (previous antibiotic treatment, UTI and hospital visit) <1 month or 1-11 months before the current UTI episode. METHODS: This was a retrospective, cohort study. Children <2 years with hospital visits and a positive urine culture obtained <48 hours from hospital admission whose culture yielded Gram-negative bacteria during the years 2015-2016, were included. RESULTS: Of the overall 744 episodes, 80% were Escherichia coli. Overall ANS rates were 66% for ampicillin; 27%-29% for amoxicillin/clavulanic acid, cephalosporins (indicating extended-spectrum beta-lactamase) and trimethoprim-sulfamethoxazole; 14% for nitrofurantoin; 5%-6% for gentamicin, ciprofloxacin and piperacillin/tazobactam; and <1% for meropenem and amikacin. Higher ANS rates were associated with Bedouin (vs. Jewish) ethnicity, exposure to risk factors <1 month, and to a lesser extent during the 1-11 months before the studied UTI episode. In episodes with risk factors <1 month, ANS rates were 81% for ampicillin; 47%-58% for amoxicillin/clavulanic acid, cephalosporins and trimethoprim-sulfamethoxazole; 19% for nitrofurantoin; 12%-23% for gentamicin, ciprofloxacin and piperacillin/tazobactam; and 2% for meropenem and amikacin. CONCLUSIONS: Previous antibiotic treatment, hospital admission and UTI, especially <1 month before the current episode, and Bedouin ethnicity, were all associated with high rates of ANS. These findings are important and may assist the treating physician in choosing an adequate empiric treatment for UTI.