Yellow fever (YF) is a significant health problem in South America and Africa. Travelers to these areas require immunization. The United States, infested with Aedes aegypti mosquitoes, is at risk of introduction of this disease. There is only a single U.S. manufacturer of YF 17D vaccine, and supplies may be insufficient in an emergency. A randomized, double-blind outpatient study was conducted in 1,440 healthy individuals, half of whom received the U.S. vaccine (YF-VAX) and half the vaccine manufactured in the United Kingdom (ARILVAX). A randomly selected subset of approximately 310 individuals in each treatment group was tested for YF neutralizing antibodies 30 days after vaccination. The primary efficacy endpoint was the proportion of individuals who developed a log neutralization index (LNI) of 0.7 or higher. Seroconversion occurred in 98.6% of individuals in the ARILVAX group and 99.3% of those in the YF-VAX group. Statistically, ARILVAX was equivalent to YF-VAX (P = .001). Both vaccines elicited mean antibody responses well above the minimal level (LNI 0.7) protective against wild-type YF virus. The mean LNI in the YF-VAX group was higher (2.21) than in the ARILVAX group (2.06; P = .010) possibly because of the higher dose contained in YF-VAX. Male gender, Caucasian race, and smoking were associated with higher antibody responses. Both vaccines were well tolerated. Overall, the treatment groups were comparable with respect to safety except that individuals in the ARILVAX group experienced significantly less edema, inflammation, and pain at the injection site than those in the YF-VAX group. No serious adverse events were attributable to either vaccine. YF-VAX participants (71.9%) experienced one or more nonserious adverse events than ARILVAX individuals (65.3%; P = .008). The difference was due to a higher rate of injection site reactions in the YF-VAX group. Mild systemic reactions (headache, myalgia, malaise, asthenia) occurred in roughly 10% to 30% of participants during the first few days after vaccination, with no significant difference across treatment groups. Adverse events were less frequent in individuals with preexisting immunity to YF, indicating a relationship to virus replication.