When we examine a mental image in the "mind's eye," we feel as though we are looking at a picture, even though the image we see is entirely in our mind. We use mental imagery to imagine possible configurations of objects, as when we think through packing a trunk or solving a jigsaw puzzle; to anticipate possible futures, as when we plan ahead in chess; or to represent the past, as when we reexamine the layout of a childhood playground. Indeed, evidence suggests that mental images are like pictures in many ways.

Studies of mental imagery often ask subjects to manipulate a remembered image in some way, for example, by searching for a particular feature, by rotating an object to compare it to a new one, or by scanning across an image like a map. By measuring how much time such operations take, we can tell something about how people represent images in their minds. Look at and memorize the map below.

Now, using the image of the map that you have committed to memory (NOT looking at the map above), find the color of the library. Next, count the number of windows in the school. In order to examine those two buildings, you needed to scan across the map from the library to the school. The time that it took you to move your mental eyes from one building to the other is called your mental scanning time. (Note: This is a difficult experiment to put into writing, since subjects would normally hear these instructions spoken, allowing them to commit their visual resources to studying the image rather than reading written words.) Mental scanning time from one object to another increases in proportion to the distance between the two objects, just as visual scanning time does with objects in the real world. This suggests that scenes and objects in mental images are represented in a way similar to scenes and objects in the real visual world.

In addition to distance, other properties of mental images also seem similar to visual images. People seem to "zoom in" to a mental image to see details of the image. Picture a goat standing next to a mouse. Subjects asked "does the goat have a beard" answer more quickly than subjects asked "does the mouse have whiskers." Because the mouse is tiny in comparison to the goat, subjects must zoom in, like a camera lens, to see the mouse's whiskers, while the goat's beard is already visible.

In another experiment, subjects were then shown two shapes at a time and asked to decide whether the shapes were the same or different. Some of the shapes were simply rotated versions of each other (e.g. upside down, or turned clockwise 90 degrees), so that the subject would need to rotate the objects in order to accurately compare them. The further the subject needed to rotate the objects, the longer it took (i.e. time to say "same" or "different" was longer for a 180-degree rotation than for a 90-degree one). This suggests that the operation of mental rotation is similar to the physical rotation of an object in space.

More than just being similar processes, visual perception and mental imagery seem to share physical and mental resources. Competition for mental resources is indicated by interference between two processes, as discussed in the section on interference. If mental imagery and visual perception (sight) use the same resources, then they should interfere with each other, causing worse performance when trying to do both at once. In an experiment designed to investigate this possibility, subjects were asked to hold a mental or auditory image in mind while trying at the same time to detect faint visual stimuli. Subjects detected fewer of the stimuli (using their vision) when trying to hold in mind the mental image, suggesting that the mental imagery interfered with their sight to some extent.

In addition to evidence from interference, evidence from brain imaging studies (see the Neuropsychology section on methodology for more information on brain imaging techniques) shows that many of the areas of the brain used for vision are activated when people perform a task requiring mental imagery. Final evidence for overlapping systems comes from people who have suffered brain damage. People who lose some visual ability as a result of brain damage also lose the corresponding ability in mental imagery; for example, people who lose the ability to see color also lose the ability to imagine scenes in color. Thus, it seems that people use the same physical areas of the brain for mental imagery and visual perception.

However, despite all their similarities, mental images are not exactly the same as visual perceptions of the real world. Mental images are influenced by our memories and other mental processes. Ambiguous pictures, like the one below, cannot be reinterpreted once they are entered into memory and become mental images.

If we first see the image as a rabbit, our mental image will show only a rabbit; we will never be able to see it as a duck by reexamining our mental image (and vice versa). In addition, our memories for images are subject to errors, such as changes due to false suggestions. There is no physical aspect to a mental image, no absolute stimulus held in mind. Rather, mental images are part of our subjective perceptions of the world, representing only what the world looks like to us.