Adriane Gaf furiCaTITDn experience suggests that most people can tell whether they are being looked at by another person who is about 8 m away.However, the results of past experiments, which used distances of no more than about 3 m, have implied that this cannot be done if the person looked at (Receiver) judges only by the iris-sclera 2 conf igura ti on of the person looking (Sender) .'!his is true even if eye contact is defined sirrply as identifying on-face gazes (FG?).It has been suggested that in everyday experience eye contact is accanpanied by cues other than iris position, and that these non-iris-position (NIP) cues to Receiver account for recognition at longer distances. '!he hypotheses of the present experiment are that FGs can be identified at considerably more than 3 m, without NIP cues: and that this happens because as features of Sender's lower face fall within Receiver's central vision at longer distances, Receiver is able to use them for triangulating the position of Sender's irises.'IWenty-six rren and 18 \\Unen acted as receivers with a male Sender at distances of 2 , 3 1/3, 4 2 /3, and 6 meters. It was expected that FG recognition at 6 m would occur, but would be hampered when Sender's lower face was masked.To detect the presence of NIP cues and to distinguish their effects from those of masking the lower face, a half-silvered mirror was at times placed in front of Receiver. Since Receiver was enclosed in a booth to darken his or her side of the mirror, he or she could see Sender clearly: but Sender's view of Receiver was blocked, so that he could not respond to the sight of Receiver returning his gaze by giving NIP cues to Receiver.Accuracy of FG recognition was not significantly lower in the mask or mirror conditions than in the no mask or mirror condition, and was significantly above chance expectation in all 3 three. The use of a mask did not appreciably hinder FG recognition at 6 rn, but did enhance it scxrewhat at 4 2/3 rn. 'Ihese resultsshowed that FGs were recognized at 6 rn, and that NIP cues were not a factor in the subjects' performance. It is suggested that masking enhanced perfo:rnence at 4 2/3 rn by eliminating distractions to using iris-sclera configuration, while at 6 rn it left both of Sender's eyes visible and within central vision.With ferrale subejcts, it was found that high visual acuity aided FG recognition at 6 rn when no mask was used. This is in accord with the contention that the high visual acuity of central vision enables the use of 10\ller facial features as reference points for determining iris position. There was not enough variation in visual acuity arrong the male subjects to assess this effect.