The mechanism of RIG-I activation has been widely studied over the past few years. RIG-I preferentially recognizes 59-triphosphorylated blunt ended double-stranded RNA, but it can also bind to long double-stranded RNA without 59ppp. The recognition of an agonist RNA triggers a conformational change, allowing RIG-I to become active thanks to the release of the CARD domains. The free CARDs are then accessible for poly-ubiquitination and recruitment of the adaptor mitochondrial antiviral signal protein. The precise mechanisms of RIG-I activation are still not fully understood. It has been proposed that RIG-I-mediated activation relies on RIG-I oligomerization via dimerization of RIG-I C terminal domain, multiple oligomerization sites within RIG-I, and/or RNA-mediated oligomerization. In the present study, we question the necessity of RIG-I self-oligomerization for signal induction. RIG-I oligomerization, induced by synthetic cognate RNA able to activate RIG-I and as well as activation by measles virus, was analysed by co-immunoprecipitation and a sensitive protein complementation assay. In the absence of convincing evidence of self-oligomerization our data support monomeric RIG-I as being the minimal signal transduction unit. RIG-I oligomerization was proposed to occur during activation by a RNA ligand by two groups in 2007–2008. Since then, the observation of RIG-I oligomerization has progressively become one of the landmarks of RIG-I activation, as many prominent papers in the field tend to report data supporting this idea. However, the biochemical support remains rather poor, and the rationale enigmatic. The RIG-I oligomerization concept originated from in vitro analysis by gel filtration of a mixture of pure RIG-I protein and short 59ppp-RNA. However, a significant shift of the volume of elution observed after chromatography does not necessarily indicate a linear augmentation of mass. Indeed the shape of the molecule can influence its migration properties through the reticulated gel and a conformational change occurs when RIG-I binds an agonist RNA with the tightening of the helicase around the RNA and the release of the CARDs. RIG-I oligomerization has also been observed by band shift in Blue Native Gel electrophoresis. In addition to some reliability concerns depending on the RNA source used to activate RIG-I, a band shift indicates a molecular change and does not necessarily prove oligomerization. The migration properties of a protein can be altered by a small bound RNA that is highly negatively charged and/or by its engagement into a multimolecular complex. In contrast, size-exclusion chromatography on a S200 column coupled to multi-angle laser light scattering analysis of mixtures of pure RIG-I protein with short dsRNA was compatible only with RNA/RIG-I 1:1 monomer complexes. In agreement with our observations, RIG-I and hairpin duplexes of 10, 20 or 30 base pairs with a single 59ppp end form 1:1 complexes as analysed by analytical ultracentrifugation-sedimentation velocity. Accordingly, crystal Dabrafenib structures of RIG-I bound to short RNA shows only monomeric RIG-I:RNA complexes in a 1:1 ratio. Only when dsRNA contains two 59 triphosphate ends, could RIGI:RNA complexes be observed in a 2:1 ratio.