Abstract Background The VARIVAX® Pregnancy Registry was established in 1995 to monitor pregnancy outcomes of women who received varicella vaccine (ie, VARIVAX) inadvertently while pregnant. Methods Health care providers and consumers sent voluntary reports about women who received VARIVAX 3 months before or during pregnancy. Follow-up occurred to evaluate pregnancy outcomes for birth defects. Outcomes from prospectively reported pregnancy exposures (ie, reports received before the outcome of the pregnancy was known) among varicella-zoster virus (VZV)-seronegative women were used to calculate rates and 95% confidence intervals (CIs). Results From 17 March 1995 through 16 October 2013, 1601 women were enrolled—966 prospectively—among whom there were 819 live births. Among 164 infants born to women who were VZV seronegative at the time of vaccination, no cases of congenital varicella syndrome (CVS) were identified (rate, 0 per 100, 95% CI, 0.0–2.2) and the birth prevalence of major birth defects was 4.3 per 100 liveborn infants (95% CI 1.7–8.6) with no pattern suggestive of CVS. No defects consistent with CVS were identified in any registry reports. Conclusions Data collected through the VARIVAX pregnancy registry do not support a relationship between the occurrence of CVS or major birth defects and varicella vaccine exposure during pregnancy, although the small numbers of exposures cannot rule out a low risk. VARIVAX remains contraindicated during pregnancy.