Research into human hearing has been ‘laboratory’ oriented in the sense that the test environments do not replicate most ‘real world’ situations (Middlebrooks & Green, 1991). It is difficult to see how such tests accurately represent ‘real world’ situations regarding sound source localisation, recognition, and navigation (walking/way-finding) performance. The research reported here was conducted during the development of an audio-based indoor navigation system. The question that arose was what constitutes ‘good’ attractor sounds versus ‘bad’ sounds? A series of physical experiments were developed to identify sounds that performed best from within a group of ‘real-world’ attractor sounds for navigability, localisation, recognition, and likeability. Experiments results were consistent with those of Landau, Wiener, Naghshineh, and Giusti (2005).