This paper considers the media research with children from a social epidemiological perspective. I am especially concerned with comparing the types of research designs, standard measures and modeling frameworks used in media studies to those used in (social) epidemiology and prevention research more generally. Overall, the state of research on the effect of various media on outcomes in children is robust and promises to be on par with the best epidemiological research. The principal measurement deficiency lies in the measurement of potential confounders and effect modifiers. With respect to research designs, the principal deficiencies include neglect of effect-modification by recognized social groupings, contextual influences and social interaction. Substantial practical benefits may come from the implementation and evaluation of intervention programs, especially group-randomized trials. As for analyses, the principal weakness appears to be in extending causal inference into observational field studies.