I study the effects of the Future to Discover Project, a randomized experiment in which Canadian high school students were either invited to participate in career planning workshops or were made eligible for an $8,000 college grant. By matching the experimental data to post-secondary institution records and income tax files, I am able to examine the effects of the interventions on college enrollment, graduation, and earnings in adulthood. I show that the career education intervention greatly improved students' outcomes in the long run by improving academic matching. In contrast, the college grant had no long-term monetary benefits despite increasing college enrollment, which is consistent with classical models of human capital investment in the absence of credit constraints. My findings suggest that informational frictions and behavioral obstacles–rather than financial constraints–represent the primary barrier to four-year college enrollment faced by low-income students. And that they explain a large part of the gap in four-year college enrollment between high- and low-income students.