Visual attention of 300 three- and five-year-old children viewing 15 hour-long Sesame Street programs in the presence of an audiovisual slide distractor was examined to establish the existence and nature of attentional inertia, defined as the increased tendency to continue looking at TV the longer one has been viewing. Three major analytic approaches were employed: 1) plotting the conditional probability function, p ( look . /look. ) , where t = look length; 2) seeking inertial effects between looks and between pauses by examining the relationship between adjacent look and pause lengths; and 3) seeking inertial effects across episodes by examining the relationship between look and pause lengths before and after bit and slide boundaries. Using TV attention data, slide attention data, and pauses in attention both to the TV and slides, a sufficient body of evidence was accumulated to conclude that attentional inertia does exist. With regard to the nature of attentional inertia, the consistent carry-over effects across bit and slide boundaries demonstrated that attentional inertia is not strictly episode-bound . All results revealed inert ial effects that were episode-free in nature. However, no results were obtained that eliminated the possibility that both episode-bound and episode-free inertia may be functioning simultaneously in visual attention. No age effects were found, and curvilinear functions were consistently obtained, suggesting that inertial effects do not continue indefinitely but eventually reach and maintain a plateau . The significance of the attentional inertia phenomenon to TV viewing and to behavior in general was discussed.