This study investigated the alleged remedial effects of blushing in the context of real-time interactions. Therefore, 30 pairs of prosocial individuals participated in a prisoner's dilemma ‘game’. The experiment was framed as an objective test of moral behaviour. To elicit a shameful moral transgression, one individual of each pair was instructed to select the non-habitual cheat-option on a pre-defined target trial. Supporting the idea that violation of shared rules elicits blushing, the defectors displayed a blush on the target trial. Yet, unexpectedly, there was a negative relationship between the observed blush intensity and the trustworthiness attributed to the defectors. One explanation might be that the ‘victims’ used the blush response to deduce and interpret the defector's motive. As the antecedent behaviour involved in the present context was not completely unambiguous with respect to the perpetrators' motive (e.g. innocent playing around vs. maximizing outcomes) the observers might have interpreted blushing as signaling that the situation should be interpreted as an intentional violation of a social standard. Together the available evidence suggests that only in the context of unambiguous antecedent behaviours blushing has remedial effects, whereas in ambiguous situations blushing has undesirable revealing effects. Copyright © 2002 John Wiley & Sons, Ltd.