Importance: SARS-CoV-2 vaccines are authorized for use in most age groups. The safety of SARS-CoV-2 vaccines is unknown in children younger than 5 years. Objective: To retrospectively evaluate the safety of the BNT162b2 vaccine used off-label in children younger than 5 years compared with the safety of non-SARS-CoV-2 vaccines in the same sample. Design, Setting, and Participants: This investigator-initiated retrospective cohort study included parents or caregivers who registered children for SARS-CoV-2 vaccination in outpatient care facilities in Germany. The study was performed as an authenticated online survey. A total of 19000 email addresses were contacted from vaccination registration databases between April 14 and May 9, 2022. Inclusion criteria were child age younger than 5 years at the first BNT162b2 vaccination and use of a correct authentication code to prove invitation. Exposures: Off-label BNT162b2 vaccination and on-label non-SARS-CoV-2 vaccinations. Main Outcomes and Measures: Reported short-term safety data of 1 to 3 doses of 3 to 10 μg BNT162b2 in children from birth to younger than 60 months are presented. Coprimary outcomes were the frequencies of 11 categories of symptoms after vaccination with bivariate analyses and regression models adjusting for age, sex, weight, and height. Results: The study included 7806 children (median age, 3 years [IQR, 2-4 years]; 3824 [49.0%] female) who were followed up of for a mean (SD) of 91.4 (38.8) days since first BNT162b2 vaccination (survey response rate, 41.1%). A 10-μg dosage was more frequently associated with local injection-site symptoms compared with lower dosages. In the active-comparator analysis, the probability of any symptoms (odds ratio [OR], 1.62; 95% CI, 1.43-1.84), local symptoms (OR, 1.68; 95% CI, 1.38-2.05), musculoskeletal symptoms (OR, 2.55; 95% CI, 1.32-4.94), dermatologic symptoms (OR, 2.18; 95% CI, 10.7-4.45), or otolaryngologic symptoms (OR, 6.37; 95% CI, 1.50-27.09) were modestly elevated after BNT162b2 compared with non-SARS-CoV-2 vaccines, whereas the probabilities of general symptoms (OR, 0.77; 95% CI, 0.63-0.95) and fever (OR, 0.42; 95% CI, 0.32-0.55) were lower after BNT162b2. Symptoms requiring hospitalization (n = 10) were reported only at BNT162b2 dosages above 3 μg. Conclusions and Relevance: In this cohort study, the symptoms reported after BNT162b2 administration were comparable overall to those for on-label non-SARS-CoV-2 vaccines in this cohort of children younger than 5 years. The present data may be used together with prospective licensure studies of BNT162b2 efficacy and safety and could help guide expert recommendations about BNT162b2 vaccinations in this age group..