School students (15–16 years) in six regions of Wales were recorded telling stories in their local English dialects. Some of these narratives were used as samples representing the main English dialect regions in Wales. Comparable groups of students (n = 169) and a group of teachers (n = 47) rated the audio-recorded speakers on a number of scales of affiliation, status, and Welshness. Statistical analysis of their ratings, employing cluster analysis and multidimensional scaling, made it possible to detect some of the competing or additive effects of dialect and narrative features. Judgments of “Welshness” of the speaker/narratives were grounded in the regional dialect properties; but other judgments, such as the likability of the speakers, tended to draw on features of both dialect and narrative. In addition, comparison of students and teachers revealed differences in their evaluations of particular dialect communities and the characteristics of the narratives. The findings illustrate the importance of approaching the analysis of dialect variation within the broader context of speech and discourse performance.