Background Universal 2-dose varicella vaccination was recommended in 2006 to further reduce varicella disease burden. This study examined 2-dose varicella vaccine effectiveness (VE) and rash severity in the setting of school-associated varicella outbreaks. Methods A case control study was conducted from January 2010 to May 2011 in all West Virginia public schools. Clinically diagnosed cases from varicella outbreaks were matched with classmate controls. Vaccination information was collected from school, health department and healthcare provider immunization information systems. Results Among the 133 cases and 365 controls enrolled, VE against all varicella was 83.2% [95% confidence interval (CI): 69.2%–90.8%] for 1-dose of varicella vaccine and 93.9% (95% CI: 86.9%–97.1%) for 2-dose; the incremental VE (2-dose vs. 1-dose) was 63.6% (95% CI: 32.6%–80.3%). In preventing moderate/severe varicella, 1-dose varicella vaccine was 88.2% (95% CI: 72.7%– 94.9%) effective, and 2-dose vaccination was 97.5% (95% CI: 91.6%–99.2%) effective, with the incremental VE of 78.6% (95% CI: 40.9%–92.3%). One-dose VE declined along with time since vaccination (VE = 93.0%, 88.0% and 81.8% in <5, 5–9 and ≥10 years after vaccination, P = 0.001 for trend). Both 1- and 2-dose breakthrough cases had milder rash than unvaccinated cases (<50 lesion: 24.6%, 49.1% and 70.0% in unvaccinated, 1-dose and 2-dose cases, P < 0.001), and no severe disease was found in 2-dose cases. Conclusions Two-dose varicella vaccination is highly effective and confers higher protection than a 1-dose regimen. High 2-dose varicella vaccination coverage should maximize the benefits of the varicella vaccination program and further reduce varicella disease burden in the United States.