The endemic character of urinary infections (UI) in sows makes collective antimicrobial therapies via feed a routine. This, however, generates sub-doses unable to heal and contribute to the selection of antibiotic-resistant bacteria. The use of individual therapy is the most appropriate procedure to be performed on animals with UI. With this study, we aimed to evaluate the occurrence of UI in sows housed in the western region of Paraná and the efficacy and cost-benefit of individual treatment. A total of 353 females were selected from five different herds, submitted to urine collection in the final third of pregnancy by spontaneous urination method. The samples were analyzed physically and chemically with the use of reagent strips, and the presence of nitrite was a determinant for positivity for UI. The animals with UI had urine submitted to a bacteriological evaluation, were treated with parenteral medication (marbofloxacin - single dose - 8 mg/kg), and submitted to a new urine collection 24h and 48h after the first. UI was observed in 4.53% of the females evaluated (16/353). Escherichia coli and Streptococcus sp. were the most frequently isolated agents. Seven days after the use of marbofloxacin 87.5% (14/16) of the animals were negative for UI, which demonstrates the efficacy of UI parenteral control. The diagnosis associated with individual therapy at the expense of collective medication was highly cost-effective, made it possible to drastically reduce the number of medicated animals, and was efficient in controlling UI. Thus, it is concluded that it is possible to make rational use of antibiotics by treating only sows that are proven to be positive for UI. This reduces the number of unnecessarily medicated animals and reduces the cost due to the use of antimicrobials only in sick animals.