Succinyl and 3'-substituted glutaryl betulin derivatives showed stronger anti-HIV activity and higher therapeutic index (TI) values than their dihydrobetulin counterparts, with ratios of 1.2:1 to 15:1 (cf. 7 and 15, 9 and 17, 10 and 18, 11 and 19, and 12 and 20). For various 3'-substituted glutaryl compounds, the order of anti-HIV effects, from strong to weak inhibition, was 3',3'-dimethyl, 3'-methyl, 3'-ethyl-3'-methyl, followed by 3',3'-tetramethylene glutaryl derivatives (10 > 9 > 11 > 12, 18 > 17 > 19 > 20). The most potent compound, 10, has two 3',3'-dimethylglutaryl groups and displays significant anti-HIV potency with an EC50 value of 0.000 66 microM and a TI of 21 515. Results for compounds (22 and 23) without a C-3 acyl group confirmed the importance of the C-3 acyl group to the anti-HIV effect. With 3',3'-tetramethylene glutaryl derivatives, triacyl 29 showed stronger inhibition than diacyl 12; in contrast, 3',3'-dimethylglutaryl compounds displayed opposite results. 3-Keto compounds (35 and 36) and 2,3-dihydro compounds (39 and 40) had EC50 values in the range of 4.3-10.0 microM, suggesting that A ring modification led to decreased potency. The reduced activity of amide (33 and 34), ester (41), and oxime (42) analogues suggested that the orientation and linkage of the C-3 acyl side chain play crucial roles in the potent anti-HIV activity. Finally, replacing the C-28 acyl group with a bulky non-carboxylic group produced a less potent compound (44). In the study of mechanism of action, our results indicated that fusion is not the primary target for the anti-HIV activity of 10. It appears to inhibit HIV replication at a late stage of the viral life cycle, i.e., after viral protein synthesis.