Estimates of disease burden and data on the sources of invasive postpartum group A streptococcus (GAS) infections will help guide public health action. Active, population-based surveillance was conducted in 9 regions from 1995 through 2000. A case of GAS infection was defined as isolation of GAS from a sterile site in a resident of a surveillance area who was pregnant or in the postpartum period. Census and live birth data were used to calculate rates. Eighty-seven cases of postpartum GAS infection (2.2% of 3957 invasive GAS infections) occurred at 3%-8% of hospitals annually. We estimate that 220 cases occurred annually in the United States. Two or more cases were noted during 6 months at 8 hospitals, during 1 year at 13 hospitals, and during 2 years at 16 hospitals. Cases due to identical emm types clustered more frequently than expected by chance. Although postpartum GAS infections are rare, the clustering of infections due to identical strains suggests that some invasive cases may have a common source and, therefore, may be preventable.