Introduction: While the lung microbiome in severe asthma (SA) has been characterized, there remains a lack of metagenomic studies in Asian cohorts. Methods: Patients with SA (N=70) were prospectively recruited, clinical data collated, and sputum samples subjected to deep shotgun metagenomic sequencing to evaluate lung microbiomes. Patients were followed up for two consecutive years and exacerbations requiring steroid bursts, emergency room visits and/or hospitalization recorded. Results: Median age of the SA cohort was 60 years (IQR 26-69) with asthma control test (ACT) score 18 (IQR 15-21) and female predominance (61%; n=43). Unsupervised spectral clustering of metagenomically-derived microbiomes identified two clusters: "Pseudomonas" (n=19) and "mixed" (n=51). The Pseudomonas cluster associated with a high expression of antimicrobial multidrug resistance and an increased exacerbation rate over two years follow-up (relative to the mixed microbiome cluster) (rate ratio 2.22; 95%CI 1.24-3.97; p=0.01). There were no significant differences between clusters for lung function, ACT score or antibiotic use. Conclusion: The airway metagenome in SA identifies a high-risk patient group characterized by Pseudomonas with increased exacerbation frequency. Funding: This research is supported by the Singapore Ministry of Health?s NMRC under its Clinician Scientist Award (CSA) (MOH-000710) (S.H.C).