Detecting anxiety is essential in both help-giving settings and in eve
ryday social relationships. The studies reported here represent a foll
ow-up to an earlier set of studies in which observers (raters) accurat
ely identified others' (expressers') state anxiety using either video-
only cues or audio-video cues. The earlier studies had included expres
sers who were repressors and nonrepressors, while the present studies
included only nonrepressors. Observers viewed/heard selected segments
of previously videotaped expressers' self-reported high and low state
anxiety experiences. Surprisingly, unlike the earlier studies, state a
nxiety was significantly inaccurately evaluated when video-only cues w
ere used, but when audio cues were available, state anxiety was correc
tly identified. Information in video-only cues seemed to mislead obser
vers since low compared with high state anxiety was rated as more anxi
ous. These studies, and the significant comparison between studies wit
h audio cues present and with audio cues absent, indicate the impact o
f auditory cues in detecting state anxiety in others.