The study examined the relationship between the number of concurrent child care arrangements and children's incidence of communicable illnesses throughout the first 4.5 years of life, and whether this association is mediated by the total number of children across care settings. Within-child fixed effects regression models were used to relate changes in the numbers of concurrent non-parental arrangements to changes in children's illnesses using longitudinal data from the NICHD's Study of Early Child Care and Youth Development (N = 1,265). 52 % of children attended multiple child care arrangements at least once from 3 to 54 months. Increases in the number of arrangements were associated with a 15 % increase in respiratory problems among children 3-54 months of age, and a 25 % increase in otitis media among children 36-54 months. Associations were smaller among African American children compared to European American and other-race children. Findings suggest that the number of peers with which a child comes into contact at child care mediates the association between increases in number of arrangements and increases in reported respiratory problems. Children attending multiple child care arrangements prior to kindergarten entry experience slightly more contemporaneous communicable diseases, relative to attending single nonparental arrangements, through exposure to more peers. (author abstract)