A nested case-control study was conducted among workers at five U.S. nuclear facilities to evaluate leukemia mortality risk (excluding chronic lymphocytic) from ionizing radiation using worksite doses and adjusting for potential confounding. Conditional logistic regression was used to estimate the relative risk (RR) of exposed workers and the excess relative risk (ERR) per unit of radiation among 206 cases and 823 age-matched controls. Adjusting for sex and benzene, the RR of leukemia for workers receiving more than 10 mSv was higher compared to those receiving lower or no dose; however, the risk increase was attenuated in the highest dose group. The ERR per 10 mSv was 1.44% (95% CI: < -1.03%, 7.59%) but was higher for workers born after 1921 compared to workers born earlier or when excluding leukemias of uncertain type. Excluding the 7% who were high-dose workers (> 100 mSv), the sex- and benzene-adjusted ERR per 10 mSv was 6.82% (95% CI: -2.87%, 24.1%). The results suggest that risks among these nuclear workers are comparable to those observed in high-dose populations, although no evidence was observed of a positive quadratic dose-response term in this study. This large study is among the first to jointly evaluate benzene and ionizing radiation risk.