In the present experiment we examined whether eye blinks could discriminate between guilty and innocent examinees in a Guilty Knowledge Test (GKT). Based on the assumption that guilty examinees would experience more cognitive load when responding to the key items than to the control items, we hypothesized that they would display fewer eye blinks during the key items than during the control items. For innocent examinees, responding to the key and control items should be equally demanding, and no differences in eye blinks between key and control items were expected. A total of 26 participants took part in an experiment where 13 guilty examinees committed a mock theft (of an exam paper) and the remaining 13 innocent examinees went on with their normal business. All participants underwent a GKT whereby their eye blinks were measured. Both guilty and innocent examinees displayed the pattern of eye blinks that we predicted. The implications of the findings are discussed.