Can adjunct faculty’s characteristics from their jobs before college employment predict their effects as college instructors? This paper is the first to link adjunct faculty’s employment history before they started teaching in college with college transcripts of students they teach. The author estimates valued-added (VA) for college faculty using a two-way fixed effects model controlling for both course- and instructor- fixed effects, then examines to what extent instructor characteristics from previous jobs (e.g., earnings) and industry experiences outside education, can predict their VA estimates. The author uses students’ future performance to infer current instructor quality. To capture multiple aspects of instructor effects, the author measures student outcomes using subsequent course performance in the subject area, credential completion, and employment and earnings. The author finds that adjunct faculty who are better at improving course performance are not always the ones who boost students’ long-term outcomes including credential completion and earnings. The author also finds that adjunct faculty who have worked in non-education sectors and had higher earnings before teaching at colleges are associated with higher VA measured by students’ long-term outcomes.