The efficacy of erythromycin stearate (ES) and its 2'-acetyl ester (erythromycin acistrate, EA) was compared in eight experimental infections in mice of both sexes. In two studies the mice were made leukopenic by whole-body irradiation. Four absorption studies were also performed in parallel. In Streptococcus pneumoniae peritonitis, the protective dose 50% (PD50) value of EA and ES, given s.c., did not differ from each other. The bioavailability of EA was slightly inferior to that of ES. In three other peritonitis studies (2 Staphylococci and 1 Streptococcus), where the treatments were given s.c., EA seemed to lag behind ES in efficacy. The parallel absorption experiment showed, however, that, as compared to ES, only about one-half of EA was released from the s.c. injection site to the blood. The adjusted PD50 values of both erythromycins were about the same, with one exception. When the treatments were given i.p. both erythromycins were equally effective, and the difference in bioavailability was minor. On the contrary, the efficacy of 2'-ethylsuccinyl erythromycin was only about one-tenth that of the other erythromycins as was also the bioavailability. Oral treatment gave similar therapeutic results with EA and ES, with similar bioavailabilities, too. In the muscle abscess model, single s.c. injections of EA and ES were equally effective in reducing the growth of Staphylococcus aureus. These results suggest that there is no great difference in the in vivo antibacterial performance of ES and its 2'-acetyl ester, although the absorption problems complicate the interpretation. Hence EA performs better than expected if only the hydrolyzed drug were useful.