An important aspect of CD8+ T cell immunity to viral infection is the complete recruitment of specific T cells into the response through activation and proliferation. One might guess that larger precursor populations would naturally dominate the immune response and previous studies have highlighted precursor frequency as a determinant of immunodominance however this is not the case during the influenza specific response in B6 mice where we see the largest and most effective responses derived from smaller naïve epitope specific precursor populations. To investigate the various factors that may dictate this inverse relationship we used immunisation experiments to equalise the abundance of various epitopes, and saw no evidence of a role for antigen levels in determining the immunodominance hierarchy after infection. We also investigated the thymic peptide repertoire which may be mediating negative selection and differentially influencing the size of epitope specific naïve precursor pools. Interestingly, analysis of CD8+ T cell avidity in both the naïve and immune populations demonstrated a correlation between the prevalence of high avidity cells within each epitope specific population and the ensuing response magnitude. In addition, we have multiple lines of evidence indicating that, within epitope-specific populations, high avidity CTL clones are preferentially expanded in both the primary and recall immune responses. Together this data suggests that the quality of the response depends on the T cell intrinsic features of epitope recognition across and within the various epitope specific populations responding to influenza infection.