The feasibility of achieving buccal delivery of morphine using the prodrug approach was assessed by studies of bioactivation, in vitro permeation and in vivo absorption. The bioactivation of various morphine-3-esters was studied in human plasma and saliva. The in vitro permeation of morphine and various morphine-3-esters was studied using porcine buccal mucosa mounted in Ussing chambers and finally the in vivo absorption of the compounds evaluated following administration in rats. Both the results from the in vitro permeation and from the absorption of the prodrugs studied suggested a parabolic relationship to the lipophilicity of the compounds. In the in vitro studies the optimal permeation was achieved for the prodrug morphine-3-propionate having a log P value of approximately 0.7. In contrast to that optimal in vivo absorption was obtained for the prodrug morphine-3-acetate having a log P value of 0.2. This discrepancy could however be explained by the enzymatic stability of the two esters in saliva, since it was found that morphine-3-propionate was more rapidly hydrolysed in saliva than was morphine-3-acetate. The study demonstrates that the buccal delivery of morphine can be markedly Improved by using ester prodrugs with higher lipophilicity than morphine itself. However, the enzymatic stability of the prodrugs in saliva also play an important role for the overall improvement in absorption properties.