BACKGROUND: A vaccine that prevents pulmonary tuberculosis in adults is needed to halt transmission in endemic regions. This trial aimed to assess the safety and immunogenicity of three administrations at varying doses of antigen and adjuvant of an investigational vaccine (ID93 + GLA-SE) compared with placebo in previously BCG-vaccinated healthy adults in a tuberculosis endemic country. METHODS: In this randomised, double-blind, placebo-controlled phase 1 trial, we enrolled HIV-negative, previously BCG-vaccinated adults (aged 18-50 years), with no evidence of previous or current tuberculosis disease, from among community volunteers in the Worcester region of Western Cape, South Africa. Participants were randomly assigned to receive varying doses of ID93 + GLA-SE or saline placebo at day 0, day 28, and day 112. Enrolment into each cohort was sequential. Cohort 1 participants were Mycobacterium tuberculosis uninfected (as defined by negative QuantiFERON [QFT] status), and received 10 μg ID93 plus 2 μg GLA-SE, or placebo; in cohorts 2-4, QFT-negative or positive participants received escalating doses of vaccine or placebo. Cohort 2 received 2 μg ID93 plus 2 μg GLA-SE; cohort 3 received 10 μg ID93 plus 2 μg GLA-SE; and cohort 4 received 10 μg ID93 plus 5 μg GLA-SE. Dose cohort allocation was sequential; randomisation within a cohort was according to a randomly-generated sequence (3 to 1 in cohort 1, 5 to 1 in cohorts 2-4). The primary endpoint was safety of ID93 + GLA-SE as defined by solicited and unsolicited adverse events up to 28 days after each study injection and serious adverse events for the duration of the study. Specific immune responses were measured by intracellular cytokine staining, flow cytometry, and ELISA. All analyses were done according to intention to treat, with additional per-protocol analyses for immunogenicity outcomes. This trial is registered with ClinicalTrials.gov, number NCT01927159. FINDINGS: Between Aug 30, 2013, and Sept 4, 2014, 227 individuals consented to participate; 213 were screened (three participants were not included as study number was already met and 11 withdrew consent before screening occurred, mostly due to relocation or demands of employment). 66 healthy, HIV-negative adults were randomly allocated to receive the vaccine (n=54) or placebo (n=12). All study participants received day 0 and day 28 study injections; five participants did not receive an injection on day 112. ID93 + GLA-SE was well tolerated; no severe or serious vaccine-related adverse events were recorded. Vaccine dose did not affect frequency or severity of adverse events, but mild injection site adverse events and flu-like symptoms were common in M tuberculosis-infected participants compared with uninfected participants. Vaccination induced durable antigen-specific IgG and Th1 cellular responses, which peaked after two administrations. Vaccine dose did not affect magnitude, kinetics, or profile of antibody and cellular responses. Earlier boosting and greater T-cell differentiation and effector-like profiles were seen in M tuberculosis-infected than in uninfected vaccinees. INTERPRETATION: Escalating doses of ID93 + GLA-SE induced similar antigen-specific CD4-positive T cell and humoral responses, with an acceptable safety profile in BCG-immunised, M tuberculosis-infected individuals. The T-cell differentiation profiles in M tuberculosis-infected vaccinees suggest priming through natural infection. While cohort sample sizes in this phase 1 trial were small and results should be interpreted in context, these data support efficacy testing of two administrations of the lowest (2 μg) ID93 vaccine dose in tuberculosis endemic populations. FUNDING: Aeras and the Paul G Allen Family Foundation.