Introduction:The Pathways to Underachievement model (PUM; Snyder & Linnenbrink-Garcia, 2013) is a person-oriented framework identifying two distinct patterns of self-and value beliefs that help explain underachievement among cognitively highly able students. This framework was developed to clear up inconclusive indings from variable-centered underachievement research, but has not been previously empirically tested. Methods: The present study aimed to test the PUM by investigating (a) whether the predicted motivational proiles are evident among a sample of high-ability students (IQ ≥ 120) beginning secondary school in Flanders, Belgium (N = 403, M age = 12.2 years, 60.5% males) and (b) whether these proiles relate to students' (dis)engagement from and (under)achievement in school, as assessed by the students, their parents and teachers, and school grades. Latent proile analysis was performed on ive motivational dimensions: academic self-concept, self-worth contingency, task value beliefs, entity beliefs, and attainment/utility value, and outcomes were compared across proiles using the BCH method. Results: Latent proile analysis identiied four proiles, each involving a distinct patterns of motivational variables. Two of the proiles exhibited maladaptive variable patterns consistent with the distinct proiles theorized by the PUM. Furthermore, proile differences in (dis) engagement and (under)achievement outcomes generally corresponded with the predictions of the PUM across multiple perspectives. Conclusion: This study found distinct maladaptive motivational proiles that were each linked to the outcomes hypothesized by the PUM. These indings establish empirically that there are qualitatively different types of high-ability underachievers, which supports the Pathways to Underachievement model and gives momentum to person-oriented analysis within underachievement research.