Matrix (i.e grid pattern) arrays from visual search studies (e.g. Flykt, in press) were presented before the probe in two dot-probe experiments. This was done to investigate to what extent probe reaction times (RTs) would reflect the RTs to targets in visual search tasks. The participants were instructed to try to ignore the arrays and to focus on fast and accurate responses to the dot-probe. The results suggest two different processing stages, at 45 and 90 ms array exposure respectively. These processing stages were interpreted as one that orients attention towards a deviant item in the array and another, latter stage, that is dependent on what overarching categories of the deviant items and the remaining items in the array belongs to. It is further suggested that caution should be taken when making inference about the mechanism behind RTs in relation to emotional stimuli.