Dual enrollment (DE) offers students the opportunity to earn college credits while in high school. Although variation in partnership contexts likely contributes to differences in DE access, structures, and outcomes, the extent of variation across DE partnerships is unclear. To build an understanding of variation in key DE structures and outcomes across partnerships, we examine DE partnerships between high schools and community colleges in Texas, where DE students account for one fifth of all community college enrollees. We use statewide administrative data from Texas to identify all DE course enrollments across three cohorts of high school students, constructing partnership-level measures and outcomes for each unique high school-college pairing. Through descriptive and regression analyses, we describe DE partnerships and estimate which DE structures and contexts predict aggregate DE course completion, college enrollment, and degree attainment. This study illuminates considerable variation in DE course structures and student composition across partnerships. Our regression results indicate that DE course structures (e.g., course subject, location, instructor type) are less practically meaningful in predicting a partnership's aggregate outcomes compared with contextual measures such as geographic locale (i.e., rural or urban) or use of an early college high school (ECHS) model. Future research should examine how to meet the needs of DE students within their geographic contexts and whether (and which) practices from ECHS models are effective for students taking DE courses outside of the ECHS model.