Undergoing an admission process (an initiation) can induce exaggerated feelings about a group, but there is little research about the role of rewards. This study replicated Aronson and Mills’ (1959) experiment. Seventy participants underwent either a severe initiation or a mild initiation. After the initiation, about half the sample received an extrinsic reward for merely completing the task. The remaining half did not receive an extrinsic reward. This was to vary the amount of dissonance. Initiation severity and reward condition had significant, noncrossing interactions. A reward led to higher group identity than no reward, supporting Levine and Moreland's (1994) group socialization model. A severe initiation did not lead to more group identity than a mild initiation; therefore, Aronson and Mills’ findings were not replicated. Interestingly, a mild initiation followed by a reward led to more group identity than a severe initiation followed by a reward. Another unexpected finding was that the extrinsic reward made no difference to group identity if the initiation was severe. Effects on ratings of the discussion were nonsignificant. Future research needs to establish how new group members ponder the severity of the admission process during the cost–benefit calculation preceding their identification with a group.