This paper is a survey of greedy randomized adaptive search procedures

(GRASP). GRASP is a multi-start or iterative procedure where each GRASP iteration con-

sists of a construction phase, where a feasible solution is constructed, followed by a local

search procedure that finds a locally optimal solution. The construction phase of GRASP

is essentially a randomized greedy algorithm. Repeated applications of the construction

procedure yields diverse starting solutions for the local search. We review a basic GRASP,

followed by enhancements to the basic procedure. We conclude by surveying operations

research and industrial applications of GRASP.