OBJECTIVES: To determine physician-administered influenza vaccine coverage for children aged 6 to 23 months in a jurisdiction with a universal influenza immunization program during 2002–2009 and to describe predictors of vaccination. METHODS: By using hospital records, we identified all infants born alive in Ontario hospitals from April 2002 through March 2008. Immunization status was ascertained by linkage to physician billing data. Children were categorized as fully, partially, or not immunized depending on the number and timing of vaccines administered. Generalized linear mixed models determined the association between immunization status and infant, physician, and maternal characteristics. RESULTS: Influenza immunization was low for the first influenza season of the study period (1% fully immunized during the 2002–2003 season), increased for the following 3 seasons (7% to 9%), but then declined (4% to 6% fully immunized during the 2006–2007 to 2008–2009 seasons). Children with chronic conditions or low birth weight were more likely to be immunized. Maternal influenza immunization (adjusted odds ratio 4.31; 95% confidence interval 4.21–4.40), having a pediatrician as the primary care practitioner (adjusted odds ratio 1.85; 95% confidence interval 1.68–2.04), high visit rates, and better continuity of care were all significantly associated with full immunization, whereas measures of social disadvantage were associated with nonimmunization. Low birth weight infants discharged from neonatal care in the winter were more likely to be immunized. CONCLUSIONS: Influenza vaccine coverage among children aged 6 to 23 months in Ontario is low, despite a universal vaccination program and high primary care visit rates. Interventions to improve coverage should target both physicians and families.