With an increasing number of everyday operations and communications becoming both automated and autonomous, ambient intelligent soundscapes are transforming to accommodate additional sonic feedback, and with it, new frameworks of listening. While this type of research and design of audio augmented technology is not new, the impact pre-existing acoustic environments upon listeners’ sense-making activities is rarely considered holistically. Much of the study into the design of effective auditory displays focuses on perceptual acuity and correct source identification, often at the expense of understanding the context of meaning-making. This paper presents a study involving 70 participants who listened to unidentified audio recordings of two archetypal everyday urban sound environments naturally containing artificial signals as well as typical sounds. Using a ThinkAloud protocol we investigated listeners’ approaches to meaning-making in both semantic and temporal dimensions. Through a semantic content analysis, we articulate five aspects of sonic meaning-making: spatial, descriptive, experiential, associational and narrative. We further analyse the use of these perceptual elements on a temporal plane, in order to investigate how listeners construct a narrative of what they hear in real-time, naturally evolving as each subsequent sound event is interpreted. Results suggest that while listeners attend to sound events and spatial characteristics of a sound environment at the beginning of a new listening situation, as the soundscape unfolds they utilize associations and familiarity in order to place individual sounds into increasingly coherent narratives. Finally, we suggest that this approach could provide sound designers and human–computer interaction specialists with a model for investigating the context aspects of a soundscape more holistically, allowing them to evaluate the effect of any new designed sounds prior to introduction into real-world environments.