The current study responds to the call for increased dialogue among different areas of additional language research. Specifically, we bring together learner corpus research and variationist approaches to second language acquisition to advance learner corpus research in two ways: (a) by modeling interlanguage development and variability and (b) by conducting an analysis that moves away from descriptive error analysis. To accomplish this goal, we analyzed grammatical gender marking in additional language Spanish. The data come from LANGSNAP, a longitudinal learner corpus. Through the use of a generalized linear mixed-effects model (a multivariate, quantitative analysis), we revealed new information about the development and variable use of gender marking in additional language Spanish. Through our analyses, we hope to contribute to greater interactions among researchers working in different strands of additional language research.