Painters, photographers, and other graphic artists continually face the problem of how to compose their subjects within rectangular frames in aesthetically pleasing ways. Despite the importance and generality of such issues, they have seldom been addressed empirically. We report the results of a series of experiments that investigate people's aesthetic responses to simple pictures consisting of one or two objects differing only in spatial position (both horizontal and vertical), size, and facing direction relative to a surrounding rectangular frame. The results, using paintings, photographs, and rendered digital images show strong, consistent preferences in the spatial composition of images depicting familiar objects, configurations of multiple objects, and even novel shapes. Viewers prefer single objects to be positioned at or near the center of the frame (the “center bias”), to face toward the center of the frame (the “facing bias”), and to occupy roughly half of the frame without cropping (the “full frame bias”). Similar preferences characterize configurations of multiple objects. The results are discussed in terms of aesthetic preferences reflecting conditions providing optimal perception of the content of the image. The implications for understanding the principles underlying art are also considered.