This paper discusses the utilization of multi-dimensional (MD) analysis (Biber, 1988; 1995) to examine linguistic variation in learner speech represented by transcribed interviews from the Louvain International Database of Spoken English Interlanguage (LINDSEI). LINDSEI is the first large-scale corpus of spoken learner English with sub-corpora of interview responses from eleven different mother-tongue backgrounds. The primary goals of this study are: to extract and identify the linguistic dimensions of learner speech from LINDSEI, to functionally interpret these dimensions, and to compare how these dimensions are distributed across speakers' eleven first language backgrounds. Results show that the four primary functional dimensions of learner speech are: Involved Conversational Style vs. Informational Production; Complex Statement of Opinion; Formal, Academic Focus of Discussion vs. Informal, Non-Academic Discourse; and Personal Narrative Prose vs. Non-Narrative Discourse. Interesting differences are observed in how these dimensions are used by learners across first language backgrounds and interview tasks.