The watching eyes effect influences people to behave more prosocially or less antisocially. It is currently unknown how the eyes themselves induce this effect. To fill this gap, we varied emotional expressions (angry, fear, happy, neutral), age (old, young), and sex (male, female) of the watching eyes to investigate which characteristics influence antisocial behaviors. Participants rated the extent they would utter 36 stereotypical statements about race, gender, and religion topics. Each statement was presented together with a pair of eyes. Approval of stereotype statements with neutral eyes did not differ from control flowers. However, age, sex, and emotional expressions of the watching eyes influenced the extent participants would utter the statements: angry old male eyes increased stereotypical statements, as did happy young eyes (males and females). Fearful eyes decreased the extent to utter stereotypical statements, except for old female eyes. Our data suggest that studies presenting eyes that appear neutral may be why replication of this effect was unsuccessful. Furthermore, the stylized eyes used in many watching eyes experiments could be interpreted as angry old male eyes to elicit changes in making donations or rating scales. We recommend further research to examine the eyes used to better understand the underlying mechanism of this effect. Collectively, our findings show that even when testing for antisocial behaviors, like the extent to utter stereotypical statements, the watching eyes effect may be better replicated with angry old male eyes or young eyes, and across different emotional expressions, age groups, and sex of the eyes.