This paper presents an application of Propensity Score Matching (PSM) to evaluations of two merit-based fellowship programs of the National Science Foundation (NSF). Evaluating such programs poses an inherent challenge: to what should differences in post-program outcomes of awardees and non-awardees be attributed? To program participation itself, or simply to the pre-existing differences in awardees’ and non-awardees’ characteristics? We address this challenge in evaluations of NSF’s International Research Fellowship Program (IRFP), an international postdoctoral fellowship for recent U.S. doctorates, and its East Asia and Pacific Summer Institutes (EAPSI) program, a summer research fellowship for U.S. graduate students. For each program, we applied PSM to estimate the impacts of program participation on subsequent international engagement and professional outcomes. A key prerequisite was data about applicants’ qualifications and experiences before award decisions were made, including, critically, each applicant’s proposal ratings by reviewers.#ResearchonEvaluation #MixedMethodsEvaluation #EvaluationUse #QuantitativeMethods-TheoryandDesign #AssessmentinHigherEducation #2012Conference #STEMEducationandTraining