The descriptive focus of conversation analysis (CA) has not been considered optimal for second language (L2) acquisition research. Recently, however, some CA researchers have addressed the developmental agenda by examining longitudinal data (e.g., Brouwer & Wagner, ; Ishida, ; Markee, ; Pekarek‐Doehler, ). The present article offers conceptual rationale and analytic demonstrations as to how CA's descriptive analyses contribute to L2 acquisition research with particular focus on English spoken discourse. While a great deal of prior L2 acquisition studies center on changes in L2 forms and functions, CA's emphasis on the sequential organization of spoken language can specify the process by which such developmental changes are occasioned in situated context of L2 use. What matters here is not just the presence of linguistic changes, but the contingent methods L2 speakers deploy in contextually occasioned language use. We present two sets of analyses in order to illustrate the case in point: cross‐sectional data for story‐prefacing work in an ESL setting, and longitudinal data on topic‐shift in an EFL setting.