In developing self-administered interviewing systems that go beyond text, survey designers are faced with choices about the virtual ―interviewer‘s‖ (VI‘s) attributes. In ACASI and IVR, for example, designers must select interviewers‘ voices. Any visual representation of an interviewer (e.g., a photograph, a video) requires choices about the interviewer‘s appearance. In earlier work we have demonstrated that such choices can affect answers; for example, we have shown that respondents report different numbers of lifetime sexual partners depending on the extent of a virtual interviewer‘s facial movement, much as occurs in comparing answers between human and ACASI administration. Here we investigate whether a VI‘s appearance--in particular the attributes that might lead respondents to think of the VI as belonging to a particular racial group or a particular gender--affects respondents‘ answers to race- and gender- related questions. In a web survey of 1735 respondents (half Black and half White, half female and half male), respondents answered (clicking or typing) questions about race and gender issues asked by one of 16 animated interviewers (Black, White, female, male). We observed race-of-VI effects on several racerelated questions, but no gender-of-VI effects on gender-related questions. For example, more respondents reported strong (rather than not strong) opposition to preferences in hiring when the question was posed by a VI identifiable as White. When respondents were asked (post-interview) to choose a VI for a hypothetical future interview, black respondents were particularly likely to choose a black VI. The fact that non-human ―interviewers‖ can elicit effects akin to those found with human interviewers raises questions about the mechanisms underlying race- and gender- of interviewer effects more generally.