This study examined predictors of support for censorship of television violence, including third-person effects (the belief that others are more affected by televised violence than oneself) and exposure to specific news stories about the issue. A random sample of 253 residents in a small, Midwestern metropolitan area participated via telephone interviews. The third-person effect for aggression predicted greater support for censorship, but the third-person effect for mean-world perceptions did not. However, when perceived effects on self and others were examined separately, greater support for censorship was associated with respondents' beliefs that violent content increased others' aggressive tendencies but their own mean-world perceptions. More frequent exposure to news stories about imitation was associated with more willingness to censor violent content, but four other types of news stories were unrelated to censorship support. Interpretations and implications of these and other findings are discussed.