In Hong Kong, universal varicella vaccination was introduced in July 2014 with a two-dose schedule but the vaccines had been available in the private market since 1996. With data from varicella notification and surveys on immunization coverage, we used the screening method to estimate dose-specific varicella vaccine effectiveness (VE) among preschool children in Hong Kong before universal vaccination. We estimated the VE of one- and two-dose varicella vaccination against all notified varicella as 69.4% (95% confidence interval (95% CI) 69.5-71.2) and 93.4% (95% CI 91.7-94.7), respectively. We found that VE did not decrease with time since receipt. Varicella vaccine was more effective against complications (85.4% [95% CI 48.8-95.8] for one dose and 100% [95% CI -Inf to 100] for two doses) and against hospital admission (75.2% [95% CI 53.4-86.8] for one dose and 93.1% [95% CI 47.1-99.1] for two doses). Lower protection of one-dose varicella vaccine resulted in breakthrough varicella. Under universal vaccination, second-dose varicella vaccine (given as combined measles, mumps, rubella and varicella vaccine) was first scheduled for children when they reach primary one (about 6 years of age) and was recently advanced to 18 months of age. Shortening the interval between the first dose and second dose of varicella vaccination should reduce breakthrough varicella and outbreaks in preschool.