Urinary tract infections (UTIs) are highly prevalent bacterial infections that pose significant health risks. Specific probiotic strains have been recommended for UTI control and management of antibiotic resistance. Otherwise, para-probiotics, defined as inactivated probiotic cells, offer potential advantages by minimizing risks associated with live microorganisms. However, the effectiveness of heat-killed probiotic strains against UTIs remains uncertain. Additionally, lactoferrin (LF), an iron-binding glycoprotein, exhibits immunomodulatory, antimicrobial, and anti-inflammatory properties. Recently, we had developed recombinant LF-expression probiotics, which can display considerate antibacterial activities against select food-borne pathogens in vitro. Thus, the present study aimed to evaluate the antibacterial activities of heat-killed natural and recombinant LF-expressing probiotics against UTIs in vitro and in vivo. Firstly, using in vitro assays, we assessed the antibacterial activity of heat-killed natural and recombinant LF-expressing probiotics against uropathogenic Escherichia coli and Klebsiella pneumoniae. Among the tested probiotics, 10 heat-killed LF-expressing strains displayed superior antibacterial efficacy compared to 12 natural probiotics. Based on their potent in vitro activity, selected probiotics were formulated into three probiotic mixtures: viable probiotic mixture (LAB), heat-killed probiotic mixture (HK-LAB), and heat-killed LF-expressing probiotic mixture (HK-LAB/LF). To further evaluate the therapeutic potential of these probiotic mixtures in vivo, we established a murine model of UTIs by intraurethral administration of E. coli to 40 female C57BL/6JNarl mice on day 0. Subsequently, mice received oral gavage of placebo, LAB, HK-LAB, or HK-LAB/LF for 21 consecutive days (n = 8 per group). An additional control group (n = 8) received ampicillin treatment for 7 days. To assess protective effects against re-infection or UTI relapse, all mice were challenged with E. coli on day 22 and E. coli plus K. pneumoniae on day 25. Results from the murine UTI model demonstrated that placebo administration did not reduce bacteriuria throughout the experiment. Conversely, supplementation with ampicillin, HK-LAB/LF, HK-LAB, or LAB significantly (p < 0.05) reduced daily bacteriuria by 103 to 104-fold on days 1, 3, 5, and 14, respectively. Furthermore, all four therapeutic treatments improved the bacteriological cure rate (BCR) with varying levels of efficacy. For the 7-day treatment course, the BCR was 25% (placebo), 62.5% (ampicillin), 37.5% (LAB), 37.5% (HK-LAB), and 62.5% (HK-LAB/LF). For the 21-day treatment course, the BCR was 25% (placebo), 75% (ampicillin), 37.5% (LAB), 37.5% (HK-LAB), and 75% (HK-LAB/LF). Notably, HK-LAB and HK-LAB/LF demonstrated superior therapeutic efficacy compared to viable LAB in treating UTIs. Overall, regarding BCR, the three probiotic mixtures can provide benefits against UTI in mice, but ampicillin therapy remains the most efficient among the four treatments. Furthermore, there was no significant difference between pre- and post-challenge courses for the two instances of re-challenging uropathogens in all mice groups, as bacteriuria levels remained below 103 CFU/mL, implying that adaptive responses of mice may help reduce the risk of recurrent UTIs. In conclusion, our results provide new evidence that oral administration of heat-killed probiotic mixtures can confer significant therapeutic efficacy against UTIs in a murine model.