The recent introduction of rotavirus vaccines into the national immunization schedules of the United States and other countries has led to substantial reductions in the incidence of severe diarrhea in children. However, there is concern over whether indirect protection evident in high-income countries in the short term will extend to low-income countries and to the long term, and whether the selective pressures imposed by vaccines will lead to the emergence of non-vaccine-type strains. Using data-driven models for the transmission dynamics of rotavirus, we generate predictions about rotavirus dynamics in response to vaccination by relating individual-level protection offered by vaccines to population-level effects. I will discuss how models were able to predict the post-vaccination emergence of a biennial pattern of epidemics in the US, why rotavirus genotypes tend to cycle, and possible explanations for the lower vaccine effectiveness observed in countries such as Malawi.