Previous studies of the incidence of congenital rubella syndrome (CRS) after rubella outbreaks have been limited because most women with infection during the first trimester elected to have their pregnancies terminated. After a rubella outbreak in 1991 we measured prospectively the impact of maternal infection on CRS among the Amish in one county in Pennsylvania. We compared rubella serology of Amish women delivering before and after the outbreak and cord blood rubella IgM from Amish and non-Amish infants. Before the outbreak 20% of Amish women were susceptible to rubella; after the outbreak 4% were (P = 0.001). Of Amish infants 15% tested positive for rubella IgM; no non-Amish infants did (P < 0.001). This rubella outbreak in a largely unimmunized community led to a high rate of CRS. The annual CRS rate among the Amish was 2130/100,000 live births. Health care providers should promote immunization in all clients and intensify efforts among the Amish.