The majority of the almost 400 confirmed cases of infant botulism in the United States have occurred in California, Pennsylvania, and Utah. In Pennsylvania, 44 of 53 (83%) cases occurred within a geographic area of Southeastern Pennsylvania which represents one tenth of the Commonwealth's area and one third of the population at risk for infant botulism. In Southeastern Pennsylvania, a map of the residences of cases circumscribes a discrete ring around Philadelphia. A case-control study performed to seek host-related risk factors, identifies the significant associations of botulism with infants who are white, breast-fed, and born at term into two-parent families with hospitalization insurance. County control studies were performed to identify differences in host-related factors between areas of high and low prevalence of botulism. Although some "protection" could be afforded Philadelphia infants by their feeding and family characteristics, the differences in case rates between Philadelphia and the botulism "ring counties" cannot be explained entirely by host-related factors. Further, the absence of botulism in counties just outside of the botulism "ring," where infants were found to have identical potential risk factors, suggests that an uneven distribution of botulinal spores in the environment is the most significant determinant of case rate.