The golden section (GS) was investigated in three experiments ( N = 91, 87, and 73 psychology students, respectively), using both traditional methods (line bi-section, production of rectangles), and novel stimuli (contours and cutouts of vases constructed by the GS and non-GS principles) and tasks (the placement of “vases” on an imaginary and a laboratory, purpose-built, mantelpiece). In five different tasks, which varied considerably in technical details, there was absolutely no evidence for the significance of the GS, nor was there a general preference for the GS vases. Instead, the search for balance seemed to motivate the subjects' mantelpiece placement choices, guided by the area (“weight”), rather than the shape, of the vases. In addition, the results cast serious doubt on the generalizability of conclusions based on the research on rectangles to real-world aesthetic objects and choices. Other substantive and methodological issues, especially with regard to the future research on the GS, and to ecological validity, were discussed.