Bactericidal antibiotic activity is necessary for cure of bacterial endocarditis. The animal model of this disease has proved extremely useful for determining the in vivo significance of in vitro observations of antibiotic effect. A review and correlation of data from in vitro, animal model, and clinical studies regarding β-lactam therapy of streptococcal and staphylococcal endocarditis are presented. Although most in vitro findings have predicted therapeutic response to infection in experimental animals and humans, discrepancies, such as the tolerance phenomenon in Staphylococcus aureus and the efficacy of certain β-lactam-aminoglycoside combinations in enterococcal infections, do exist. In general, where examples exist, the results in animal models correlate with results from clinical studies of endocarditis.