The possibility has been raised in recent years that infection might contribute as an inflammatory stimulus to chronic "noninfectious" degenerative diseases. Within the past decade, serious attention has been given to the possibility of bacterial vectors as causal factors of atherosclerosis. To date, the greatest amount of information has related to the intracellular organism Chlamydia pneumoniae. This interest has been stimulated by the frequent finding of bacterial antigens and, occasionally, recoverable organisms, within human atherosclerotic plaque. Indirect evidence for and against the benefit of anti-Chlamydia antibiotic agents comes from epidemiologic studies. Given the potential for confounding in observational studies, prospective, randomized intervention trials are required. These antibiotic trials have generated enthusiastic expectations for proving (or disproving) the infectious-disease hypothesis of atherosclerosis and establishing new therapies. However, these expectations have been tempered by important limitations and uncertainties. Negative outcomes can be explained not only by an incorrect hypothesis but also by inadequate study size or design or by an ineffective antibiotic regimen. In contrast, if studies are positive, the hypothesis still is not entirely proved, because a nonspecific anti-inflammatory effect or an anti-infective action against other organisms might be operative. The clinical trial data to date have not provided adequate support for the clinical use of antibiotics in primary or secondary prevention of coronary heart disease. New and innovative experimental approaches, in addition to traditionally designed antibiotic trials, should be welcome in our attempts to gain adequate insight into the role of infection in atherosclerosis and its therapy.