The main finding of this thesis is that rhythm is a stratified variable in the speech of men in Stockholm. An epicenter for the social forces associated with late modernity, Stockholm is also home to Europe’s ‘first’ multiethnolect (Rinkeby Swedish, Kotsinas 1988a). Swedish-language researchers describe the variety as ‘staccato’, but rhythm has not been thoroughly investigated for any variety of Stockholm Swedish in production. Data come from 36 male Stockholmers, ages 24–45, from a stratified sample of social classes. Seventeen self-identify by the term svensk (Swedish) and 19 by the highly racialized term invandrare (literal translation: immigrant). All were born in Sweden save for three who arrived before age seven. Three contextual styles were elicited to capture a speech-formality cline: CASUAL, READING, and RADIO (reading like a radio announcer). Rhythm is operationalized with an adaptation of the nPVIV algorithm (Low, Grabe, & Nolan 2000). Not only does rhythm stratify predictably in the direction of staccato (low alternation) for the racialized working class, it also is significantly high-alternation/non-staccato in the speech of the white working class. The former is interpreted to be a feature of multiethnolect; the latter a feature of Södersnack, Stockholm’s industrial-era working-class variety. The higher classes produce an intermediate degree of rhythm in their casual speech. Rhythm variation among the working classes is also stylistically sensitive. Working-class READING and RADIO appear to target upper-class CASUAL. The racialized working class shows a stylistic sensitivity that is stronger among younger speakers than old, implying a transition from indicator to marker (Labov 1972a:179) for staccato rhythm. The white working class shows a high degree of stylistic sensitivity regardless of age, implying that high alternation is a Södersnack legacy feature. Generational differences in rhythm production are examined within the racialized working class, and a change point is identified between those born before 1983 and those after. Those born before 1983 mainly achieve ‘staccato’ with a reduction of accented phonologically-long vowels. Those born after 1983 achieve it with an innovation; they enlarge unstressed vowels, both phonologically short and long. ‘Reduction’ and ‘enlargement’ refer to duration, f0, and energy. The change point coincides with historical spikes in migration, inequality, and school segregation that would have occurred when the speakers were in adolescence. In all contextual styles, age is a stable predictor of rhythm, independent of social class and racialization. Younger speakers of any background have more staccato speech than older speakers of the same background. It is proposed that this is due to the diffusion of contact prosody, for which multiethnolect is one key conduit.