We conducted two studies to examine how interviewers' nonverbal behaviors affect children's perceptions and suggestibility. In the first study, 42 8- to 10-year-olds watched video clips showing an interviewer displaying combinations of supportive and nonsupportive nonverbal behaviors and were asked to rate the interviewer on six attributes (e.g., friendliness, strictness). Smiling received high ratings on the positive attributes (i.e., friendly, helpful, and sincere), and fidgeting received high ratings on the negative attributes (i.e., strict, bored, and stressed). For the second study, 86 8- to 10-year-olds participated in a learning activity about the vocal chords. One week later, they were interviewed individually about the activity by an interviewer adopting either the supportive (i.e., smiling) or nonsupportive (i.e., fidgeting) behavior. Children questioned by the nonsupportive interviewer were less accurate and more likely to falsely report having been touched than were those questioned by the supportive interviewer. Children questioned by the supportive interviewer were also more likely to say that they did not know an answer than were children questioned by the nonsupportive interviewer. Participants in both conditions gave more correct answers to questions about central, as opposed to peripheral, details of the activity. Implications of these findings for the appropriate interviewing of child witnesses are discussed.