Rising Lyme borreliosis incidence rates, potential for severe outcomes, and limitations in accurate and timely diagnosis for treatment initiation suggest the need for a preventive vaccine; however, no vaccine is currently available for human use. We performed two studies in adults to optimise the dose level and vaccination schedule for VLA15, an investigational Lyme borreliosis vaccine targeting outer surface protein A (OspA) serotypes 1-6, which are associated with the most common pathogenic Borrelia species in Europe and North America. Both randomised, observer-blind, placebo-controlled, multicentre phase 2 studies included participants aged 18-65 years without recent history of Lyme borreliosis or tick bites. Study one was conducted at nine clinical research and study centre sites in the USA (n=6), Germany (n=2), and Belgium (n=1); study two was conducted at five of the study one US sites. Based on a randomisation list created by an unmasked statistician for each study, participants were randomly assigned via an electronic case report form randomisation module to receive 90 μg (study one only), 135 μg, or 180 μg VLA15 or placebo by intramuscular injection at months 0, 1, and 2 (study one) or 0, 2, and 6 (study two). Study one began with a run-in phase to confirm safety, after which the Data Safety Monitoring Board recommended the removal of the 90 μg group and continuation of the study. In the study one run-in phase, randomisation was stratified by study site, whereas in the study one main phase and in study two, randomisation was stratified by study site, age group, and baseline B burgdorferi (sensu lato) serostatus. All individuals were masked, other than staff involved in randomisation, vaccine preparation or administration, or safety data monitoring. The primary endpoint for both studies was OspA-specific IgG geometric mean titres (GMTs) at 1 month after the third vaccination and was evaluated in the per-protocol population. Safety endpoints were evaluated in the safety population: all participants who received at least one vaccination. Both studies are registered at ClinicalTrials.gov (study one NCT03769194 and study two NCT03970733) and are completed. For study one, 573 participants were screened and randomly assigned to treatment groups between Dec 21, 2018, and Sept, 26, 2019. For study two, 248 participants were screened and randomly assigned between June 26 and Sept 3, 2019. In study one, 29 participants were assigned to receive 90 μg VLA15, 215 to 135 μg, 205 to 180 μg, and 124 to placebo. In study two, 97 participants were assigned to receive 135 μg VLA15, 100 to 180 μg, and 51 to placebo. At 1 month after the third vaccination (ie, month 3), OspA-specific IgG GMTs in study one ranged from 74·3 (serotype 1; 95% CI 46·4-119·0) to 267·4 units per mL (serotype 3; 194·8-367·1) for 90 μg VLA15, 101·9 (serotype 1; 87·1-119·4) to 283·2 units per mL (serotype 3; 248·2-323·1) for 135 μg, and 115·8 (serotype 1; 98·8-135·7) to 308·6 units per mL (serotype 3; 266·8-356·8) for 180 μg. In study two, ranges at 1 month after the third vaccination (ie, month 7) were 278·5 (serotype 1; 214·9-361·0) to 545·2 units per mL (serotype 2; 431·8-688·4) for 135 μg VLA15 and 274·7 (serotype 1; 209·4-360·4) to 596·8 units per mL (serotype 3; 471·9-754·8) for 180 μg. Relative to placebo, the VLA15 groups had more frequent reports of solicited local adverse events (study one: 94%, 95% CI 91-96 vs 26%, 19-34; study two: 96%, 93-98 vs 35%, 24-49 after any vaccination) and solicited systemic adverse events (study one: 69%, 65-73 vs 43%, 34-52; study two: 74%, 67-80 vs 51%, 38-64); most were mild or moderate. In study one, unsolicited adverse events were reported by 52% (48-57) of participants in the VLA15 groups and 52% (43-60) of those in the placebo groups; for study two these were 65% (58-71) and 69% (55-80), respectively. Percentages of participants reporting serious unsolicited adverse events (study one: 2%, 1-4; study two: 4%, 2-7) and adverse events of special interest (study one: 1%, 0-2; study two: 1%, 0-3) were low across all groups. A single severe, possibly related unsolicited adverse event was reported (worsening of pre-existing ventricular extrasystoles, which resolved after change of relevant concomitant medication); no related serious adverse events or deaths were reported. VLA15 was safe, well tolerated, and elicited robust antibody responses to all six OspA serotypes. These findings support further clinical development of VLA15 using the 180 μg dose and 0-2-6-month schedule, which was associated with the greatest immune responses. Valneva.