Vaccine efficacy largely depends upon DC targeting and activation. The most potent TLR soluble ligands induce diffuse DC activation, which may be associated with marked pro-inflammatory responses and possibly adverse effects. This raises the concern that effective vaccine adjuvants may similarly rely on widespread DC activation. Using a promising candidate vaccine against tuberculosis (fusion protein of Ag85B and 6-kDa early secretory antigenic target (ESAT-6)) formulated in the potent IC31 adjuvant, DC targeting and activation was studied in vivo, following the fate of antigen and adjuvant in the draining lymph nodes, to define the magnitude of DC targeting/activation required in vivo to induce protective vaccine responses. Unexpectedly, protective IFN-gamma-mediated Ag85B-ESAT-6/IC31 responses were associated to the activation of a minute population (less than 0.3%) of CD11c(+) lymph node DC, without detectable systemic pro-inflammatory responses. This activated peripheral tissue-derived DC population, characterized by enhanced CD80, CD86, CD40 and IL-12p40 expression, was only identified when focusing on adjuvant- or antigen-labeled CD11c(+) DC, which were found to support T cell proliferation. Immunization with aluminum hydroxide adjuvant (Alum) resulted in a similar proportion of antigen-associated DC but without detectable enhancement of CD80, CD86, CD40 or IL-12p40 expression. Thus, potent protective IFN-gamma-producing responses may be elicited by the exquisite activation of a minute number of in vivo targeted DC.