Anxiety affects a great proportion of the population worldwide, mostly females, with profound effects on quality of life and increased disease risk. The diet-gut-brain axis has emerged as a promising preventive and treatment avenue, yet the key determinants and contextual factors (e.g., habitual diet, enterotypes) remain unclear. This thesis combined systematic literature research, experimental and clinical methods, metagenomics analysis and computational approaches to disentangle how diet and gut microbiome features relate to anxiety severity in young women with subclinical anxiety. A critical literature evaluation first identified theoretical and methodological research gaps and informed the study design. A cross-over experimental study then compared a food recording app (Libro) with a 24-hour recall method (Intake24) observing suboptimal performance of Libro and supporting Intake24 for subsequent short-term dietary assessment. Cross-sectional and longitudinal analysis investigated how long- and short-term diet, gut microbiome composition, functional capacity, and microbial metabolite production relate to anxiety, accounting for context-dependent effects in a cohort of 46 females (18–24 years). Two participant clusters were identified based on habitual diet quality: lower (HEI 1) and higher (HEI 2). HEI 2 reported lower anxiety than HEI 1 and a more stable diet quality over time. Only in HEI 2 anxiety covaried with diet quality and fat intake. Across clusters, vegetable and ultra-processed food intake showed the largest between-subject contribution to anxiety. Several gut microbial species and pathways were linearly and non-linearly associated with anxiety. The Bacteroides-enriched enterotype showed a more pro-inflammatory gut profile, lower vegetable intake, and the least favourable diet-gut-anxiety associations. Finally, microbial ursodiol production was associated with lower anxiety and attenuated the positive link between ultra-processed food intake and anxiety. Overall, this thesis highlights promising targets for pro-, pre-, postbiotic and dietary interventions, and emphasizes the importance of context-dependent mechanisms to inform personalized intervention trials aimed at reducing anxiety.