Evidence of the risk of birth defects with efavirenz use is limited. We updated a meta-analysis of birth defects in infants with first trimester efavirenz exposure up to July 2011. In 21 studies, there were 39 defects among live births in 1437 women receiving first trimester efavirenz [2.0%, 95% confidence interval (CI) 0.82-3.18]. The relative risk of defects comparing women on efavirenz-based (1290 live births) and nonefavirenz-based regimens (8122 live births) was 0.85 (95% CI 0.61-1.20). One neural tube defect was observed (myelomeningocele), giving an incidence of 0.07% (95% CI 0.002-0.39).