This study tested the hypothesis that elementary visuo-motor functions involved in visual scanning, as measured by fixation and saccadic tasks, are better in a group of high-level clay target shooters (N=7) than in a control group (N=8). In the fixation task, subject were told to keep fixation as still as possible on a target for 1 min, both in the presence and absence of distracters. For shooters, time did not have an effect on fixation stability, and they had more stable fixation than controls in the distracters condition. Results indicate a difference between groups on both the temporal span of attention and selective attention. In the saccadic task, subjects were asked to saccade, as fast as possible, towards a peripherally displayed target. Two conditions were used: simple reaction to target onset and discrimination between targets and distracters. Shooters had faster saccadic latency to targets than controls in both conditions. Finally, to evaluate the effect of exercise on saccadic latency, we trained one control subject to saccade to a target displayed at a constant spatial position. At the end of the training, saccadic latency reached a value comparable to that recorded in shooters. Learning was largely retinotopic, not showing transfer to untrained spatial positions.