In this research, we investigate the effect of group identity, amount of money, and type of action on engagement in corrupt behavior. We conducted a 2 x 2 x 2 experiment in which participants responded to scenarios where they stated how likely they were to engage in corruption. In these scenarios, group identity (ingroup or outgroup), the amount of money at stake (high or low), and type of action (passive or active) were manipulated to evaluate their influence on this ethical decision-making process. We also tested past dishonest behavior and social norms (injunctive and descriptive ones) as predictors of corruption intentions. Our main results are: (i) participants who admitted to having engaged in past dishonesty more often reported stronger intentions of corrupt behavior; (ii) descriptive norms were significant predictors of corruption intentions; (iii) individuals tended to engage in corruption more often when dealing with ingroup members and lower amounts of money. These results indicate that individuals who engage in petty dishonesty may be more corruptible. Our findings also suggest the relevance of investing in public policies to make norms of honesty salient in order to prevent corruption and that, to increase effectiveness, such policies should consider the role of group identity, impeding that negotiations be held among ingroup members.