Primary and secondary allylic acetates and bentoates react with the dimethyl (pheny1)silyl -cuprate reagent to give allylsilanes, provided that the THF in which the cuprate is prepared is diluted with ether before addition of the allylic ester. The reaction is reasonably regioselective in some cases: (i) when the allylic system is more-substituted at one end than the other, as in the reactions 4-5 and 9 --+ 10; (ii) when the steric hindrance at one end is neopentyl-like, as in the reactions 15 -+ 16;and (iii) when the disubstituted double bond has the 2 configuration, as in the reactions Z-19-+ E-21 or, better, because the silyl group is becoming attached to the less-sterically hindered end of the allylic system, Z-20-+ 2-22. The regioselectivity is better if a phenyl carbamate is used in place of the ester, and a three-step protocol assembling the mixed cuprate on the leaving group is used, as in the reactions 23 -+ 24 and E-or 2-29 + E-21, or, best of all, because the silyl group is again becoming attached to the less-sterically hindered end of the allylic system, E-or 2-30-E-22. This sequence works well t o move the silyl group onto the more substituted end of an ally1 system, but only when the move is from a secondary allylic carbamate t o a tertiary allylsilane, as in the reaction 38 + 39. Allyl(trimethyl)silanes can be made using alkyl-or aryl-cuprates on trimethylsilyl-containing allylic esters and carbamates, as in the reactions 40 -+ 41, and 43 + 44. The reaction of the silylcuprate with allylic esters and the three-step sequence with the allylic carbamates are stereochemically complementary, the former being stereospecifically anti and the latter stereospecifically syn. Homochiral allylsilanes can be made by these methods with high levels of stereospecificity, as shown by the synthesis of the allylsilanes 54,58 and 59.Allylsilanes undergo electrophilic substitution reactions regiospecifically in the SE2' sense,' and stereospecifically in an anti sense, as shown, principally, by the work of Wetter,' Es~henmoser,~ Kumada? Kitching and ourselves.6 To take full advantage of the highly stereospecific and regiospecific reactions of allylsilanes, and anticipating a need to test whether the osmylation, epoxidation, Simmons-Smith r e a~t i o n ,~ hydroboration * and protodesilylation ' of allylsilanes, reported in the three preceding papers to this one, might be similarly well controlled, we sought new ways of making allylsilanes. There are, of course, many ways of making allylsilanes," but it was imperative that the new routes should be both regiocontrolled and stereocontrolled, for which there were no general methods. In this paper and that following, we describe two such routes, the first using allylic alcohols as substrates, and the second using diastereoselective aldol reactions followed by stereospecific decarboxylative eliminations. ' We have reported some of the present work using allylic alcohol derivatives in two preliminary communications. 'We had already established13 a simple synthesis of allylsilan...