A meta-analysis of six controlled studies was performed to evaluate the risk of cytomegalovirus (CMV) infection among pediatric nurses. The pooled risk ratio for CMV infection in pediatric nurses using cumulative incidence data was statistically significant (risk ratio [RR] 2.7; 95% confidence interval [CI] 1.33 - 5.52), but person-year analysis taking account of follow-up periods demonstrated a trend toward increased risk that failed to reach statistical significance (RR 1.8; 95% CI 0.88 - 3.55). Despite pooling, there was low statistical power for comparing person-year rates. The studies included in this analysis failed to provide data on several potential confounding variables. We conclude that studies published prior to the widespread adoption of universal precautions suggest that pediatric nurses may have been at increased risk for CMV infection due to occupational exposure, but inadequate design and sample size of the studies prevent a definitive conclusion. Well-designed, controlled studies are still needed to define the occupational risk of CMV infection.