OBJECTIVE: Our objective was to assess whether the occurrence of medically attended local reactions to intramuscularly administered vaccines varies by injection site (arm versus thigh) in children 1 to 6 years of age. METHODS: This is a retrospective cohort study of children in the Vaccine Safety Datalink population from 2002 to 2009. Site of injection and the outcome of medically attended local reactions were identified from administrative data. RESULTS: The study cohort of 1.4 million children received 6.0 million intramuscular (IM) vaccines during the study period. The primary analyses evaluated the IM vaccines most commonly administered alone, which included inactivated influenza, hepatitis A, and diphtheria-tetanus-acellular pertussis (DTaP) vaccines. For inactivated influenza and hepatitis A vaccines, local reactions were relatively uncommon, and there was no difference in risk of these events with arm versus thigh injections. The rate of local reactions after DTaP vaccines was higher, and vaccination in the arm was associated with a significantly greater risk of this outcome compared with vaccination in the thigh, both for children 12 to 35 months (relative risk: 1.88 [95% confidence interval: 1.34-2.65]) and 3 to 6 years of age (relative risk: 1.41 [95% confidence interval: 0.84-2.34]), although this difference was not statistically significant in the older age group. CONCLUSIONS: Injection in the thigh is associated with a significantly lower risk of a medically attended local reaction to a DTaP vaccination among children 12 to 35 months of age, supporting current recommendations to administer IM vaccinations in the thigh for children younger than 3 years of age.