The paper presents a simple game theoretic model in which players decide on the search procedures for a prize which is located in one of a labeled set of boxes. The prize is awarded to the player who finds it first. A player can decide on the number of (costly) search units he employs and the order in which he conducts the search. It is shown that in equilibrium, the players employ an equal number of search units and conduct a fully random search. The paper demonstrates that the search procedure is intrinsically inefficient in two senses: the players employ a non-optimal number of search units and they may open the same empty box twice.