Bacterial vaginosis (BV) is the most common reason for abnormal vaginal discharge in reproductive-age women and one of its most important causative agents is the gram-variable bacterium Gardnerella vaginalis. BV is not accompanied by significant local inflammation, whereas the "fishy odor" test is always positive. In contrast, aerobic vaginitis (AV) is predominantly associated with Escherichia coli, but Streptococcus agalactiae and Staphylococcus aureus are also involved. Standard treatment of BV consists of oral or intravaginal antibiotics, although these are unable to spontaneously restore normal flora characterized by a high concentration of lactobacilli. The main limitation is the inability to offer a long-term defensive barrier, thus facilitating relapses and recurrences. This study was undertaken firstly to assess the ability of selected lactobacilli to in vitro antagonize G. vaginalis to determine an association with a strain able to inhibit E. coli, thus identifying a possible use in AV. The second step of the study was to conduct a human pilot trial in women affected by BV using an association of the most promising and active bacteria. For this purpose, neutralized supernatants of individual lactobacilli were tested at percentages ranging from 0.5% to 4% to determine their ability to hinder the growth of G. vaginalis American Type Culture Collection 10231. The bacterium that was able to exert the strongest inhibition was subsequently tested with Lactobacillus plantarum LP01 in a human intervention, placebo-controlled, pilot trial involving 34 female subjects (aged between 18 and 50, mean 34.7±8.9, no menopausal women) diagnosed with BV. The 2 microorganisms Lactobacillus fermentum LF15 (DSM 26955) and L. plantarum LP01 (LMG P-21021) were delivered to the vagina by means of slow-release vaginal tablets, also containing 50 mg of tara gum. The amount of each strain was 400 million live cells per dose. The women were instructed to apply a vaginal tablet once a day for 7 consecutive nights, followed by 1 tablet every 3 nights for a further 3-week application (acute phase) and, finally, 1 tablet per week to maintain a long-term vaginal colonization against possible recurrences. A clinical examination was performed and the Nugent score was quantified for each patient at enrollment (d0), after 28 days (d28), and at the end of the second month of relapse prevention (d56). A statistical comparison was made between d28, or d56, and d0, and between d56 and d28 to quantify the efficacy against possible recurrences. L. fermentum LF15 showed the strongest in vitro inhibitory activity towards G. vaginalis American Type Culture Collection (ATCC) 10231 after both 24 and 48 hours. In the human trial, the 2 lactobacilli selected, namely L. fermentum LF15 and L. plantarum LP01, significantly reduced the Nugent score below the threshold of 7 after 28 days in 22 patients of 24 in the active group (91.7%, P<0.001). Eight women (33.3%) recorded a Nugent score between 4 and 6, evidence of an intermediate situation, whereas the remaining 14 (58.3%) showed a score <4, therefore suggesting the restoration of physiological vaginal microbiota. At the end of the second month, only 4 women registered a Nugent score >7, definable as BV (16.7%, P=0.065 compared with d28). In the placebo group, no significant differences were recorded at any time. BV, also known as vaginal bacteriosis is the most common cause of vaginal infection in women of childbearing age. Furthermore, BV is often asymptomatic as about 50% of women with this condition have no symptoms at all and the prevalence rate in apparently healthy women is around 10%. This study suggests the ability of the 2 strains L. fermentum LF15 and L. plantarum LP01 to counteract acute Gardnerella infections effectively and significantly improve the related uncomfortable symptoms in a very high percentage of women. This could be partially attributed to the presence of tara gum, which is able to create a mechanical barrier against Gardnerella on the surface of vaginal mucosa as a primary mechanism. Furthermore, long-term physiological protection seems to be established, thanks to the integration of the 2 lactobacilli into the vaginal microbiota and to their adhesion to the epithelial cells of the mucosa. In the light of the additional in vitro inhibitory activity against E. coli, their prospective use in AV could also prove interesting.