State-sponsored merit-based scholarship programs have proliferated since the 1990s. Using student-level data across institutions, we estimate the effects of merit-based scholarships on first -time students’ characteristics and their first-year academic behaviors, with specific focus on engineering students. Difference-in-differences regression models are used in the analysis. We find that merit-based scholarships did not help improve SAT score or first-year GPA of resident students including those majoring in engineering. The average socioeconomic indicator of engineering students increased in the presence of merit-based scholarships, suggesting the socioeconomic diversity of engineering experienced a decline after scholarship adoption. Institutional differences existed in the mean SAT score and socioeconomic status, and this stratification was more prominent in engineering. The presence of merit-based scholarships was associated with an increased likelihood that resident students would withdraw from courses and take summer courses after scholarships were implemented. The effects of merit-based scholarships on first-year course load were disproportionate across states, following credit hour requirements in scholarship retention policies. Compared to total residents, engineering residents were more prone to reduce first-year credit hours with a merit-based scholarship in effect,possibly indicating that coursework for engineering students required more effort so that students chose a lighter load to maintain good grades.