High-quality evaluations of both efficacy and safety are essential to characterize the risk-benefit profile of antibiotics. The current US Food and Drug Administration guidelines on trial design for community-acquired pneumonia have several weaknesses, including the failure to insist on the use of double blinding and intention-to-treat analysis. A primary difficulty with noninferiority designs is that poorly conducted studies, which increase noise, are biased toward the conclusion of noninferiority even in the presence of important differences between the test drug and the control drug. Additionally, results of noninferiority trials, in the absence of a well-established anchor, may be difficult to interpret. The US Food and Drug Administration drug-evaluation process includes preclinical studies to assess toxicity and a series of clinical studies involving humans to define efficacy and to identify potential safety problems. When signals are apparent, as they were for telithromycin (liver toxicity in rats, dogs, and monkeys) or for sparfloxacin (prolongation of the QT interval in dogs), it is nonetheless essential that these signals are fully and fairly evaluated in human studies with adequate power. Risks of antibiotic use, such as prolongation of the corrected QT interval and sudden death, may be acceptable in the presence of convincing benefits for patients with severe, life-threatening infections; however, the risk-benefit profile derived from severe infections cannot necessarily be generalized to mild or self-limited infections, for which the same serious drug-associated risks are likely to exceed the benefits. Complete and proper evaluation of both safety and efficacy in specific situations is essential to define the risk-benefit profiles of all drugs, including antibiotics.