Abstract The Implicit Association Test (IAT), designed to measure how strongly an individual implicitly aligns two dichotomies, such as male/female vs. home/career, is adapted in this study to analyze sociolinguistic variation. Three experiments test the IAT's ability to detect relationships between sociolinguistic variables and social categories and between related sociolinguistic variables. In Experiment 1, associations are shown between the English variable (ING) on the one hand and Northern/Southern states, blue collar/white collar professions and country singers/news anchors on the other, using written forms. In Experiment 2, associations are shown between audio tokens of (ING) and Northern/Southern states and between (ING) and /ay/ monophthongization, both tied to the ideologically well-defined Southern accent, but not between (ING) and /t/ release, which do not share an enregistered variety. Experiment 3 documents the lack of correlation between IAT measures and two more explicit measures of association: direct question and social evaluation. These results suggest that sociolinguists should take dual processing models of social cognition into account in considering the cognitive processing of socially loaded linguistic forms and that the IAT may be useful tool for the analysis of sociolinguistic meaning.