This study verified the influence of a punishment reference histogram on the sentences of lay judges. In Experiment 1, 80 undergraduates determined the prison terms of an offender in an audio trial. About half the students were shown a histogram with a peak of the punishment meted out in similar cases and the remainder a histogram without it. The peak was set as a relatively short prison term (5–10 yrs) considering the seriousness of the crime (murder). The result indicated that the peak-histogram group chose significantly shorter prison terms than did the non-peak-histogram group, suggesting that the peak influenced the former’s sentencing. Moreover, Experiment 2 showed that the peak of the histogram influenced participant sentencing more strongly than a table, although both held the same punishment data, and that the same shaped histogram of different data did not. From these results, we concluded that the anchoring effect of the histogram is most likely to be caused by the visual attractiveness of the peak.