Sexual harassment has been associated with suicidal behaviors, and with the rise in suicides in the U.S. military, sexual harassment's role in suicide has been of growing interest. Lacking are studies that examine group- or unit-level variables in the relationship of sexual harassment to suicidal behaviors (thoughts, plans, and attempts). In this study, survey data from soldiers (12,567 soldiers in 180 company-sized units) who completed the Unit Risk Inventory administered during calendar year 2010 were analyzed using hierarchical linear modeling. At the individual level, sexual harassment was associated with a fivefold increase for risk of suicide. Reporting that leaders could be trusted was associated with a decreased suicide risk by about one-third. There was no statistically significant interaction between sexual harassment and trusted leaders in predicting the suicidal behaviors. At the group level, units or companies having higher levels of sexual harassment also had soldiers three times more at risk for suicide. A cross-leveling effect was also observed: Among units having higher levels of sexual harassment, the negative correlation (buffering effect of unit leaders on suicidal behaviors) was diminished. Implications of findings for preventing sexual harassment and suicide risk are discussed.