This study was an exploratory investigation of beauty in architecture, through which a theory of waste predicting beauty in buildings was developed. In a series of four experiments, it was shown firstly, through an experiment where a selection of University of Edinburgh buildings were rated for beauty, that buildings do differ significantly in their ratings of beauty, that functionality does not affect beauty ratings and that agreement exists between individuals on what is, and is not, attractive in architecture. After this, evolutionary theories of beauty, which cite waste as a crucial factor in creating beauty, such as peacock’s tails or birdsong, are applied to architecture. In this way it is predicted that buildings with a high waste value, observed through not maximising potential floor space or by incorporating excessive detail into the building design, will be rated as more attractive than those buildings which are less wasteful. This was investigated through a number of experiments. Firstly, when pictures from an architectural book were rated for waste by experimenters and then rated by participants for beauty, a significant positive correlation of 0.762 was found. After this an experiment was conducted in which 22 participants described their ideal house. By analysing the qualitative data recorded, it was possible to sort the descriptions into clusters. It was found that 19 of the participants described an extremely similar house and could be placed in one group, the remaining 3 descriptions were idiosyncratic in nature and so were kept separately. From the descriptions of the houses, ‘Google Sketchup’ drawings were created of the 4 houses. These were then rated on waste and beauty by 19 participants in the final of the four experiments. The hypothesis that waste ratings would influence beauty ratings was convincingly supported with a significant Pearson’s correlation of 0.857 and a R2 value of 0.739 found. The theory that waste determines beauty in architecture has therefore been supported.