Death related to acute hepatitis B occurs in approximately 1%of patients. We investigated an outbreak of hepatitis B virus (HBV) infections among injection d rug users (IDUs) resulting in several deaths. We conducted a case-control study of fulminant (case patients) and nonfulminant (control patients) HBV infections . We directly sequenced the entire HBV genome from fulminant and nonfulminant ca ses. From October 1998 to July 2000,21 acute HBV infections, including 10 fulminant hepatitis B cases, were identified. The med ian age was 30 (range, 18-49) years, 12 (57%) were female, 20 (95%) were Amer ican Indians, and 20 (95%) reported injecting illicit drugs. All patients with fulminant hepatitis B died (case-fatality rate = 47.6%). Case patients (n = 5) and control patients (n = 9) were similar with respect to age, sex, race, and h epatitis C virus serostatus. All case patients used acetaminophen during their i llness compared with 44%of control patients (P = .08). Compared with control pa tients, case patients lost more weight in the 6 months before illness (P = .04); during their illness, they used more alcohol (P = .03) and methamphetamine (P = .04). All 9 isolates sequenced were genotype D, shared 99.7%homology, and incl uded mutations previously described in association with fulminant hepatitis B. I n conclusion, a high prevalence of exposure to factors potentiating hepatic dama ge with acute hepatitis B contributed to the outbreak’s high mortality rate; mu tations present in the outbreak strain might also have been a factor. Improved v accination coverage among IDUs has the potential to prevent similar outbreaks in the future. We investigated an outbreak of hepatitis B virus (HBV) infections among injection d rug users (IDUs) resulting in several deaths. We conducted a case-control study of fulminant (case patients) and nonfulminant (control patients) HBV infections. We directly sequenced the entire HBV genome from fulminant and nonfulminant ca ses. From October 1998 to July 2000,21 acute HBV infections, including 10 fulminant hepatitis B cases, were identified. The med ian All patients with fulminant hepatitis B died (case%, 18-49) years, 12 (57%) were female, 20 (95%) were Amer ican Indians, and 20 -fatality rate = 47.6%). Case patients (n = 5) and control patients (n = 9) were similar with respect to age, sex, race, and h epatitis C virus serostatus. All case patients used acetaminophen during their i llness Compared with 44% of control patients (P = .08). Compared with control pa tients, case patients lost more weight in the 6 months before illness (P = .04); during their illness, they used more alcohol (P = .03) and methamphetamine (P = .04). All 9 isolates sequenced were genotype D, shared 99.7% homology, and incl uded mutations previously described in association with fulminant hepatitis B. I n conclusion, a high prevalence of exposure to factors potentiating hepatic dama ge with acute hepatitis B contributed to the outbreak’s high mortality rate; mu tations present in the outbreak strain might also have been a factor. Improved v accination coverage among IDUs has the potential to prevent similar outbreaks in the future.