Limited effectiveness data are available comparing live attenuated influenza vaccine (LAIV) to inactivated influenza vaccine (TIV) among adults. To compare the incidence of influenza-like illness following immunization of adults with LAIV vs. TIV, we conducted a retrospective cohort analysis of active component U.S. military personnel for the 2005-2006 and 2006-2007 influenza seasons. Recruits experienced a much higher burden of disease compared to non-recruits, with crude incidence rates of influenza-like illness 2-16 times higher than non-recruits depending on the season and cohort. For both seasons, a slightly greater protection from influenza-like illness was found for non-recruits who received TIV compared to LAIV (adjusted incidence rate ratio, 1.17 (95% CI, 1.14-1.20) and 1.33 (95% CI, 1.30-1.36), 2005-2006 and 2006-2007 influenza seasons, respectively). However, for Army and Air Force recruits, LAIV was found to provide significantly greater protection from influenza-like illnesses compared to TIV, with adjusted incidence rates of influenza-like illness 22-51% and 18-47% lower among LAIV compared to TIV recipients for the 2005-2006 and 2006-2007 influenza seasons, respectively. Possible reasons for differences in recruit and non-recruit findings include differences in pre-existing influenza antibody levels, differing respiratory disease burden, and/or unmeasured confounding. Consideration of these findings should be made when developing influenza immunization policies.