BACKGROUND Antibody response to influenza vaccine is limited in early. Infants have poorer hemagglutination-inhibiting antibody responses than 12-month-old. Intradermal administration reportedly elicited immune responses similar to or better than a standard intramuscular dose. We hypothesized that intradermal injection could achieve a better response in infants than subcutaneous injection. METHODS We randomized 34 healthy infants 6-12 months old to either intradermal immunization (0.1 ml of trivalent influenza vaccine containing at least 3 microg of hemagglutinin antigen per strain) or subcutaneous immunization (also 0.1 ml). Changes in hemagglutination inhibition titer were compared using Mann-Whitney U-test, changes in positivity rate, seroconversion, and seroprotection. Local and systemic adverse events were assessed. RESULTS All 32 infants received both injections. Antibody titers on days at 42 after intradermal injection were significantly greater than subcutaneous injection (P=0.032 in A/New Caledonia (H1N1), 0.019 in A New York (H3N2) and 0.044 in B/Shanghai. Positive titers for A New York (H3N2) were attained significantly more often after intradermal (73.3%) than subcutaneous injection (23.5%) on day 28, and significantly more often 42 days after intradermal injection (93.3% for A/New Caledonia (H1N1) and 73.3% for B/Shanghai) than after subcutaneous injection. Positive rates for other stains were similar between groups on days 28 and 42. Seroconversion rates were similar between groups. Seroprotection on day 42 for A New York (H3N2) was significantly greater in the intradermal (86.7%) than in the subcutaneous group (35.3%). Seroprotection rates for other stains were similar. CONCLUSIONS Intradermal administration to infants of two doses of influenza vaccine was more immunogenic than subcutaneous injection. Seroconversion and seroprotection rates remained insufficient. Further study of route, quantity, and frequency are needed to improve of responses in infants.