Selective attention is a key concept in communication research despite equivocal supporting evidence. This study focuses on methodological concerns in the study of selective attention. Using unobtrusive measures of selective attention to online political content this study found consistent support for the selective attention hypothesis. Three within-subjects quasi experiments were conducted (n = 29, n = 36, and n = 75) that measured subjects' attention to information consistent with and counter to their beliefs. Two other measures of selective attention were also used: (1) whether subjects turned first to consistent or counter information, and (2) how deeply into the material they read.