Individuals cheat more in groups than when alone, though the exact reasons for this are not clear. Here, we use a lab-based experiment (N = 116) to ask whether (i) variation in risk-proneness and (ii) the perception of subjective social costs, such as being perceived negatively by others or the consequences of letting others down, motivate people to behave dishonestly. In addition, we asked whether self-reported social and ethical risk preferences impact dishonesty and whether these interact with social pressure. The most striking result of this study is the very low levels of dishonesty observed: despite using established methods that have demonstrated high levels of dishonesty in previous studies, only seven (6 %) participants were dishonest in this task. It is possible that contextual factors greatly affect dishonesty, though we also discuss the possibility that previous experiments have over-estimated the extent to which people will lie for financial rewards in laboratory experiments.