Objective: This study examined whether psychological resilience mediates the association between anxiety and PA among Pakistani university students and, as a secondary aim, explored heterogeneity in these relationships using person-centred approach. Methods: A cross-sectional survey was conducted among 770 university students (aged 18–26 years) from six cities in Punjab, Pakistan, between May and December 2025. Participants completed validated measures of anxiety (GAD-7), psychological resilience (CD-RISC-10), and self-reported PA. Correlation, regression, mediation, and latent profile analyses were performed, controlling for gender and academic year. Results: Anxiety was significantly and negatively correlated with both resilience (r = −0.475, p < 0.001) and PA (r = −0.378, p < 0.001), while resilience was positively associated with PA (r = 0.419, p < 0.001). Physical activity was significantly predicted by anxiety (B = −0.383, p < 0.001) and resilience (B = 0.174, p < 0.001) in regression models, explaining 21.7% of the variance in PA. The mediation analysis revealed that resilience explained 39% of the total effect of anxiety on PA, indicating partial mediation. LPA revealed three distinct profiles: an Anxious–Inactive group (27.1%) characterised by high anxiety and low resilience; a Moderately Distressed–Moderately Active group (47.0%) with intermediate levels; and a Resilient–Active group (25.8%) with low anxiety and high resilience. Conclusions: Lower physical activity among Pakistani university students is partly explained by reduced psychological resilience in the context of anxiety. These findings highlight the importance of resilience as a protective resource and support integrated mental health and physical activity interventions in resource-limited university settings.