Attentional biases have been observed in the perception of geometric forms (Avrahami, 1998, Perception, 27: 431-438; Latimer et al., 1999, Quarterly Journal of Experimental Psychology, B in press). For example, subjects distinguish forms more quickly when their discriminating features are located at top and right rather than at other positions within the forms. This paper further explores detection of stimuli as a function of retinal locus while the eyes hold a centre-field fixation or are in smooth pursuit of a small moving target. In the static condition, subjects gazed at a central fixation point, and while holding this fixation, were required to detect a stimulus which could be a small cross or a small circle appearing randomly at one of 128 positions within the visual quadrants: top-left, top-right, bottom-left and bottom-right. The subject pressed one of two response buttons to signal the cross and the other button to signal a circle. Eye movements during trials were monitored, and trials during which a subject failed to maintain centre fixation were rejected. As hypothesised, targets appearing in the top-right quadrant were detected faster than those in other quadrants. In the smooth-pursuit condition, subjects tracked a small white dot moving within a square window. Again, small target circles or crosses appeared at random in the four visual quadrants and eye movements were monitored to ensure that subjects were pursuing the moving dot during detection of the peripheral targets. Early results suggest again that detection of targets appearing top-right of gaze point are detected faster than those appearing in other quadrants. Explanations of the attentional bias in terms of experience at reading from left to right and optic flow from top to bottom of the visual field during forward locomotion are discussed. An artificial neural network that acquires top-right bias when it is trained to recognise patterns that scroll onto its visual field from top to bottom (as in forward locomotion) and right to left (as in reading text) is described.