A13~TRACT MlIl:llillll j:-. ;111 IlllP0l'l;lll{ genclic opcr;lIioll (11:11 11l:lp~ (0 Illailliaill the genetic divCI'sity of lile poplildtioll ill order to ;lChiew a good :-.olulion lo an oplimization problem. The conven[ional mutalion, with its r;llldorn nature and low probability of occurrence, i:-. llll;lhlc [0 guide the algorithm for fast convergence. In this <.lrticle, we propose a new Illut:llion technique called, directed mutation following the concept of induced mutatioll ill hiol(Jgical systems. Directed mutation deterministically introduces new points in the pllj1ulation guided (directed) by the inl'orll1ation acquired in lhe previous gel1er;llil)lh. It expedites tile ;i1gorillllll using an acceleration function which Call he bdsed OLl ~I ;,c1lenl or extrapolation. The crf'cctivcness of the new technique has been dCll1onslr;llcd through optimization of ,I sci of complex functions, ;Ind thc results have also becn comp;lreu with those of the cOllventional GAs.