This study examined the differential impact of trait and state attachment dimensions on social and emotional loneliness among first-year undergraduates in India. Participants were 535 freshmen (Mean age = 17.8 years, SD = 0.84; 113 females) enrolled at an Indian institute. Experiences in close relationships - Revised - General short form (ECR-R-GSF) was used to measure trait anxiety and avoidance, while State adult attachment measure (SAAM) was used to assess state anxiety and avoidance of the participants. Social and emotional loneliness were measured by the 6-item De Jong Gierveld Loneliness Scale. Regression analyses revealed distinct predictive patterns for the two forms of loneliness. For social loneliness, both trait avoidance (β = .23, p < .001) and trait anxiety (β = .23, p < .001) emerged as equally strong, significant positive predictors, whereas state anxiety (β = −.04, ns) and state avoidance (β = .08, p = .076) did not exert meaningful effects. The model explained approximately 16% of variance (R² = .156). In contrast, emotional loneliness was predicted only by trait anxiety (β = .37, p < .001), with trait avoidance (β = .02, ns), state anxiety (β = .08, p = .101), and state avoidance (β = .03, ns) contributing negligibly. This model accounted for 18% of the variance (R² = .183). These findings underscore the centrality of stable trait attachment dispositions in shaping loneliness outcomes among young adults, with social loneliness jointly influenced by trait avoidance and anxiety, and emotional loneliness predominantly predicted by trait anxiety. The limited role of state attachment suggests that situational fluctuations in attachment are less consequential for loneliness in this cohort.