Objectives To assess the immunogenicity of vaccines for infants mid to investigate whether the incidence of reactogenicity is reduced after each immunisation dose using needles of varying lengths and gauges. Design Randomised controlled trial. Setting 18 general practices within two UK primary care trusts. Participants 696 healthy infants vaccinated at 2, 3, and 4 months of age, with follow-up to 5 months of age. Interventions Combined diphtheria, tetanus, whole cell pertussis, and Haemophilus influenzae type b vaccine and a serogroup C meningococcal glycoconjugate vaccine administered using either a wide, long needle (23 gauge/0.6 turn diameter, 25 mm), a narrow, short needle (25 gauge/0.5 mm diameter, 16 mm), or a narrow, long needle (25 gauge, 25 mm). Main outcome measures Local mid general reactions recorded by parents for three days after each dose; and diphtheria, tetanus, mid H influenzae type b antibody concentrations and functional antibody against serogroup C Neisseria meningitidis 28-42 days after the third dose. Results Local reactions to diphtheria, tetanus, whole cell pertussis, H influenzae type b vaccinations decreased significantly with wide, long needles compared with narrow, short needles. At all three doses one less infant experienced local reactions at days 1, 2, or 3 for every six to eight vaccinated. Significantly fewer infants vaccinated with the long needle experienced severe local reactions. Non-inferiority of the immune I response was shown using a wide, long needle rather than a narrow, short needle for serogroup C meningococcal glycoconjugate vaccine and for diphtheria but not for H influenzae type b or tetanus, although no evidence was found of a decrease. Little difference was found between needles of the same length but different gauges in local reaction or immune response. Conclusions Long (25 mm) needles for infant immunisations can significantly reduce vaccine reactogenicity at each dose while achieving comparable immunogenicity to that of short (16 mm) needles.