To test the hypothesis that use of sexist language in television newscasts cultivates images of women as immature, frivolous, or incompetent, two television news anchors (one male, one female) were asked to tape versions of a news story involving a 28-year-old female lottery winner. In one version, each anchor referred to this female as "woman," in the other they referred to her as "girl." The four experimental treatments (male vs. female broadcaster, sexist vs. nonsexist linguistic choice) were edited into four newscasts by the same anchors. The newscasts were then shown to 95 college students who were randomly assigned to one of the four treatments and who were asked to rate the lottery winner on evaluative, potency, and activity semantic differential scales. Results did not produce a generally sexist language effect. However, female subjects rated the lottery winner higher when she was referred to as "woman," whereas just the opposite was true among male subjects, who rated the female higher when referred to as "girl." Also, subjects who viewed the female anchor rated the lottery winner as more active than did those who viewed the male anchor, raising the question of anchor-related contributions to television news effects. Data tables and figures are appended. (HOD)