The experimental paradigm of our research is described in detail in the presentation of Groner et al. (Stimulus competition in saccadic programming I : Experimental technique and landing sites of the first saccades). There, with respect to the direction of attention, it was distinguished between (1) stimulus driven processing (= bottom up) where salient aspects of the stimulus attract attention, and (2) expectancy driven processing (top-down) where the observer is attending to those aspects of the stimulus which are related to his or her expectation. We investigated the relative influence of the two aspects by varying (i) the salient features of the stimulus - by varying the content of the images, spatial frequency content of the stimuli, and contrast - as well as (ii) the search intention of the observer - by instructing them to decide whether the target is within the search display or not. It is demonstrated that the saccadic eye movements are, in part, controlled by stimuli with high spatial frequency and high contrast (bottom up processing). The saccades are also controlled, in part, by the search for the target: In fact, the relatively most probable landing position of any fixation is a hit, i.e. the fixated stimulus matches the target perfectly (top down processing). Nevertheless, the interaction of both processes is not clear. The probability of a hit in the first fixation is only about 25% (expecting 12.5% at random), whereas it is higher in the second fixation. This result is in contradiction with the assumption of immediate processing, and we will try to resolve this problem by models of delayed or parallel processing. To evaluate the extent of peripheral processing, a control experiment was run where the observers were asked to fixate the centre of the screen, suppressing any saccades, and to decide whether the target was present or not.