Abstract: Japanese judges are least likely to hang a defendant for murder if they graduated from a high-status university, passed the bar-exam-equivalent quickly, or enjoy a fast-track career within the courts. “Panel composition effects” and other measures of collegiality seem unrelated to sentencing patterns. To explore the effect of judicial panel composition beyond the more-often-studied world of politically prominent cases, I examine its impact on criminal sentencing. More specifically, I examine the possible determinants of the propensity of Japanese judges to sentence guilty defendants to death. Toward this end, I collect all opinions published since 1980 in murder cases—about 200 cases. Because each case involves a three-judge panel but some judges write multiple opinions, these cases involve about 440 judges. Within this group, the most elite judges are least likely to impose the death penalty. Measures of possible collegiality—how long judges have served on a court together, graduation from a common university, closeness in age—have no observable impact. The presence of potential “whistle-blower” judges also appears not to matter.