Survey measures of attention, in addition to the usual measures of exposure to media use, are explored in detail. Using data from a two-year longitudinal study of adolescents and their parents, the reliability and validity of self-report questions about attention to newspaper and television news are examined. Attention measures proved to be highly stable across three waves of interviews; repeated items were summed to form reliable indices of attention to each medium's news, and to general news topics without specific mention of medium. Changes in knowledge about public affairs and politics were assessed to test the effects of both exposure and attention to media news. These tests showed significant increments of knowledge gain associated with media attention, even after exposure to the medium was controlled in hierarchical regression analysis. Addition of the attention measures was especially important in evaluating television news, which contributes much more to public knowledge than has been suggested by previous studies that use only measures of TV exposure.