Protein synthesis begins with recruitment of the 40S ribosomal subunit to an mRNA. In eukaryotes at least twelve translation initiation factors and an mRNA methylated G-cap are required, while in many viruses a structured RNA element, or IRES, replaces most factors and the mRNA cap. In both cases the translation factor complex eIF3 is required to coordinate formation of an initiation complex and prevent premature association of the 40S and 60S subunits. In collaboration with the Doudna lab (UC Berkeley/HHMI) we have recently obtained the cryo-EM structure of human eIF3 and characterized its interaction with Hepatitis C virus (HCV) IRES and eIF4G, the major component of the cap-binding complex (Siridechadilok et al., Science 2005). eIF3 has a mutually exclusive binding site for IRES and eIF4G, implying a mechanistic overlap that helps explains why this RNA obviates the requirement for cap-binding complex in translation initiation. Combining our own and published structural analysis of different binary complexes we have produced models for the ternary complexes 40S-eIF3-IRES and 40S-eIF3-eIF4G that show how eIF3 positions both IRES and so as to be able to anchor the attached mRNA near the exit site on the 40S ribosomal subunit (Siridechadilok et al., Science 2005). The position of the eIF3 provides a plausible explanation for how this factor inhibits premature assembly of the 40S with the 60S ribosomal subunit. In summary, this work suggests conserved interactions coordinated by eIF3 that place the mRNA in the ribosomal decoding center and prepare the 40S subunit for assembly into active ribosomes, both for the canonical initiation pathway, and for the simplified, IRES-mediated pathway. Among our immediate goals are the localization of different subunits within eIF3, and the interaction of eIF3 with other initiation factors, with the ultimate goal of a structural characterization of the full initiation complex.