Background: Determining the root causes of persistent underrepresentation of different subpopulations in engineering remains a continued challenge. Because place-based variation of resource distribution is not random and because school and community contexts influence high school outcomes, considering variation across those contexts should be paramount in broadening participation research. Purpose/Hypothesis: This study takes a macroscopic systems view of engineering enrollments to understand variation across one state's public high school rates of engineering matriculation. Design/Method: This study uses a dataset from the Virginia Longitudinal Data System that includes all students who completed high school from a Virginia public school from 2007 to 2014 (N=685,429). We explore geographic variation in four-year undergraduate engineering enrollment as a function of gender, race/ethnicity, and economically disadvantaged status. Additionally, we investigate the relationship between characteristics of the high school and community contexts and undergraduate engineering enrollment across Virginia's high schools using regression analysis. Results: Our findings illuminate inequality in enrollment in engineering programs at four-year institutions across high schools by gender, race, and socioeconomic status (and the intersections among those demographics). Different high schools have different engineering enrollment rates among students who attend four-year postsecondary institutions. We show strong associations between high schools' engineering enrollment rates and four-year institution enrollment rates as well as moderate associations for high schools' community socioeconomic status. Conclusions: Strong systemic forces need to be overcome to broaden participation in engineering. We demonstrate the insights that state longitudinal data systems can illuminate in engineering education research.