Age-related macular degeneration (AMD), a leading cause of legal blindness, causes irreversible visual impairment within a central scotoma (blind spot). Patients with AMD therefore use extrafoveal (not in the center of vision) locations for fixation, with a cost of increased fixation instability and reduced viewing efficiency. The clinical symptoms of AMD are varied. To investigate the degree to which observers with different types of AMD could retain residual visual functioning but without pathological complications, we studied eye movement (EM) patterns with a shape identification task with 4 types of simulated central scotoma (Relative, Absolute, Distorted, Warped) with young normal observers in Experiment 1. Results showed that the Relative scotoma (reduced contrast) was the least disruptive form. A complete loss or a distortion of central visual input (Absolute, Distorted, Warped) led to greater reduction of response efficiency and more severely altered EM patterns. There was a preferential horizontal fixation shift away from the targets toward the right visual field, with little shift along the vertical. Generally, there was a marked improvement with practice. Larger scotomata always led to larger changes, except for the Relative type. Since AMD affects mostly older people, we compared young and old healthy adults in Experiment 2. The results showed that old adults were less efficient than young adults, tended to have larger fixation shifts along the x- and Y-axes, but with a smaller angular change, and were slower to adapt. Large scotomata were especially disruptive. Our results suggest that when using simulated scotomata, one needs to adjust the properties of the scotoma to the kind of AMD studied. It is also preferable to use old adults, since their response appears to be different in many ways from that of young adults. This novel perspective might suggest new requirements for low vision training.