This paper investigates the effects of individual, family, social and demographic characteristics on the maintenance of Spanish among English-speaking Latino youth. This research finds effects of generation, gender, race, parent's English proficiency, single-parent status, parental income, and neighborhood concentration of co-ethnics as well as combined effects of race and gender on Spanish oral proficiency. The findings presented here suggest support for elements of the assimilation and the segmented assimilation theoretical perspectives as well as the race-gender experience theory. The author suggests that Spanish-speaking proficiency may be associated with opportunities to speak Spanish that are structured differently, not only by family and neighborhood contexts that allow for greater or lesser contact with Spanish, but also by gender and race.