Many vaccines for bioterrorism agents are investigational and therefore not available (outside of research protocol use) to all at-risk laboratory workers who have begun working with these agents as a result of increased interest in biodefense research. Illness surveillance data archived from the U.S. offensive biological warfare program (from 1943 to 1969) were reviewed to assess the impact of safety measures on disease prevention (including biosafety cabinets [BSCs]) before and after vaccine availability. Most laboratory-acquired infections from agents with higher infective doses (e.g., anthrax, glanders, and plague) were prevented with personal protective measures and safety training alone. Safety measures (including BSCs) without vaccination failed to sufficiently prevent illness from agents with lower infective doses in this high-risk research setting. Infections continued with tularemia (average 15/year), Venezuelan equine encephalitis (1.9/year), and Q fever (3.4/year) but decreased dramatically once vaccinations became available (average of 1, 0.6, and 0 infections per year, respectively). While laboratory-acquired infections are not expected to occur frequently in the current lower-risk biodefense research setting because of further improvements in biosafety equipment and changes in biosafety policies, the data help to define the inherent risks of working with the specific agents of bioterrorism. The data support the idea that research with these agents should be restricted to laboratories with experience in handling highly hazardous agents and where appropriate safety training and precautions can be implemented.