Homicide is the second leading cause of death on the job for workers in the United States. To identify workplace-level predictors of homicide risk, a case-control study of worker killings in North Carolina in 1994-1998 was conducted. Workplaces were the units of analysis: case workplaces (n = 105) were those where a worker was killed during the study period, while controls (n = 210) were a density sample of North Carolina workplaces, matched on time and industry sector. Potential risk and protective factors were assessed in telephone interviews with workplace managers. Associations were measured by the exposure odds ratio and 95% confidence interval, estimated via conditional logistic regression. Characteristics associated with notably higher risk included being at the current location for 2 years or less (odds ratio (OR) = 5.3, 95% confidence interval (CI): 2.2, 12.6), having only one worker (OR = 2.9, 95% CI: 1.2, 7.2), and having night (OR = 4.9, 95% CI: 2.7, 8.8) or Saturday (OR = 4.2, 95% CI: 1.9, 9.2) hours. Workplaces with only male employees (OR = 3.1, 95% CI: 1.5, 6.5) or with African-American or Asian employees were also more likely to experience a killing. While few of the preceding risk factors are directly modifiable through workplace interventions, it is important to identify them before developing or evaluating preventive measures.