In the present study we examined the antisaccade cost (latency difference between antisaccades and prosaccades) in a variety of search tasks. In a series of experiments participants searched for a target and were required to execute a saccade toward (prosaccade) or aw ay (antisaccade) from the target. The results revealed that the antisaccade cost was greater for static targets than for dynamic targets, and it was greater for onset targets than for offset targets. Furthermore, the offset of an onset target interfered with prosaccades, but facilitated antisaccades, resulting in a reduction of the antisaccade cost. To account for the data a model is presented, in which attentional control and working memory processes play an important role in the generation of antisaccades.