In this paper, scientific evidence for alcohol and other drug testing programs is compared to determine whether each approach is justifiable for improving work-place safety. Three types of studies are reviewed: laboratory, epidemiological and evaluation studies. Laboratory studies show that alcohol use decreases psycho-motor performance; however, for other drugs, some drugs deteriorate performance while others have little effect. Epidemiological studies in the work-place have not provided conclusive evidence that a strong causal link exists between either alcohol or other drug use and work-place injuries/accidents. Evaluation studies have not shown that either drug or alcohol testing significantly reduces work injuries/accidents. Other types of scientific evidence provide some justification of alcohol testing, but not for other drug testing. Alcohol use is much more common than other drug use in industrialized countries, suggesting that alcohol may be more related to industrial accidents than other drugs. As well, epidemiological studies suggest that alcohol is a major factor for traffic collisions, but research is inconclusive for other drugs. Finally, alcohol testing is more justifiable than drug testing because the results of alcohol tests closely correlate with psycho-motor performance while drug tests do not.