BACKGROUND Age-related macular degeneration (AMD) impairs central vision and can also impact reaching movements (Timberlake et al., 2011). When the fovea degenerates, a preferred retinal locus (PRL) in the periphery is utilized as the oculomotor reference. Delays in reach initiation may be linked to delays in saccade initiation (Renninger, Ma-Wyatt & Fletcher, 2012), however the cause remains unclear. PURPOSE Determine whether saccade delays are caused by an online re-referencing to the PRL or to delays in target localization. METHODS Maculopathy and control subjects look at and reach rapidly to an isolated visual target. The target is presented in one of 8 directions at eccentricities of 2-12 degrees. Binocular eye movements are measured with an adapted Eyelink 1000, tower mount configuration (head restrained). Reach start and endpoints are timed and measured with a mouse button release and stimulus display contact (ELO touchscreen), respectively. Accuracy and timing feedback is given. Time penalty was adjusted to enforce rapid movements. Monocular microperimetry was performed on maculopathy subjects for central scotoma mapping and PRL localization. Fovea location was obtained by an OCT scan (Optos OCT/SLO). Fixation stability was measured by the OCT/SLO and Eyelink. RESULTS Maculopathy subjects exhibit both a decrease in fixation stability and an increase in misdirected saccades as compared to control observers. Despite the misdirection, PRLs were near the target at time of finger contact. Spatial divergence of saccade endpoints and finger endpoints occur along the fovea-PRL axis, suggesting a role of both reference frames in movement generation. Vector analysis also implies that movements could be PRL-directed, fovea-directed or a directional average of the two. CONCLUSION Eye position signals continue to play a role in eye-hand movement targeting, even in the absence of foveal vision.