Every year, the federal government distributes $11 billion in education benefits to nearly one million veterans (GAO, 2013). Despite the substantial price tag and reach of these benefits, we understand very little about how veteran students fare in postsecondary programs and why outcomes may be different for veteran students. Theory and related evidence predict that veteran students should be less successful than their nonveteran peers, yet the limited past research suggests that they are actually as successful as, if not more successful than, nonveterans. This is the student veteran paradox. I posit seven potential explanations to resolve this paradox: bias in past research, background characteristics of veterans, enrollment behaviors of veterans, maturation from delayed entry, education aid benefits for veterans, unobservable factors associated with selection into the military, or the direct effects of military service. I use OLS regression and logistic regression to assess three metrics of student success: grades, retention, and completion. I also leverage variations in the GI Bill program to assess whether higher levels of funding lead to better student success outcomes. Finally, I use matching to test whether unobservable factors associated with military enlistment or the direct effects of military service could drive veteran student success. Student veterans hold many characteristics that predict lower probabilities of college success, but veterans and nonveterans generally have similar academic outcomes. When controlling for background characteristics, enrollment patterns, age, and term of entry, predicted first year GPA is lower for veterans, but veterans are more likely to return after the first year and are more likely to graduate. Generally, students with higher levels of veteran education benefits xii have better retention and graduation outcomes, but aid levels seem to have little impact on first year grades. Veterans still have lower grades than similar matched nonveterans, but the veterans are more likely to return after the first year and are more likely to graduate. For retention and graduation, these results rule out the bias, background characteristics, and maturation explanations, but support the enrollment patterns and funding explanations. The results are consistent with the direct effects explanation, but the selection explanation cannot be ruled out completely.