It is well established that voters’ ability and willingness to punish politicians that engage in corruption varies. In this paper, we argue that A potential yet overlooked explanation that voters may judge the misuse of public money differently depending on the type of policy project impacted by the misuse and the gender of the offender. Voters’ willingness to punish corruption will be influenced by whether the corrupt politician misused public funds in a policy sector that is coded in line with the politician’s gender—that is, whether a woman (man) committed corruption in a feminine-coded (masculine-coded) policy area. Furthermore, we argue that effects will differ by respondent gender, specifically we expect that female respondents will react particularly strongly to misconduct by female politicians in female coded policy areas. We test these expectations with a conjoint experiment in Mexico that tasks respondents with evaluating hypothetical mayors who embezzled money from different policy sectors. We leverage two female coded policy areas (healthcare and welfare) and two traditionally male coded policy areas (crime and infrastructure. Our findings speak to the role gender stereotypes play in punishing corruption, and provide evidence that the policy area in which corruption occurs can change voters’ willingness to electorally punish malfeasant mayors.