I study a single-agent sequential search problem as in Weitzman. Contrary to Weitzman, conditional on stopping, the agent may take any uninspected box without first inspecting its contents. This introduces a new trade-off. By taking a box without inspection, the agent saves on its inspection costs. However, by inspecting it, he may discover that its contents are lower than he anticipated. I identify sufficient conditions on the parameters of the environment under which I characterize the optimal policy. Both the order in which boxes are inspected and the stopping rule may differ from that in Weitzman’s model. Moreover, I provide additional results that partially characterize the optimal policy when these conditions fail.