BACKGROUND: Asthma, a heterogeneous disease, can be divided into 4 inflammatory phenotypes using induced sputum cell counts-eosinophilic asthma (EA), neutrophilic asthma (NA), mixed granulocytic asthma, and paucigranulocytic asthma (PGA). Although research has focused on EA and NA, there is little known about PGA. OBJECTIVE: To study the heterogeneity of PGA and identify possible PGA clusters to guide clinical treatment. METHODS: Patients with PGA were grouped by hierarchical cluster analysis and enrolled into a prospective cohort study to validate the clusters, relative to future risk of asthma exacerbations in a real-world setting. Clusters were validated by tree analysis in a separate population. Finally, we explored PGA stability. RESULTS: Cluster analysis of 145 patients with PGA identified 3 clusters: cluster 1 (n[ 110, 75.9%) was "mild PGA," cluster 2 (n [ 20, 13.8%) was "PGA with psychological dysfunction and rhinoconjunctivitis and other allergic diseases," and cluster 3 (n = 15, 10.3%) was "smoking-associated PGA." Cluster 3 had significantly increased risk of severe exacerbation (relative risk =RR] = 6.43, P =.01), emergency visit (RR = 8.61, P =.03), and hospitalization (RR = 12.94, P <.01). Results of the cluster analysis were successfully validated in an independent PGA population classified using decision tree analysis. Although PGA can transform into or develop from other phenotypes, 70% were stable over time. CONCLUSIONS: Among 3 identified PGA clusters, cluster 3 had a higher risk of severe exacerbation. PGA heterogeneity indicates the requirement of novel targeted interventions. (C) 2021 American Academy of Allergy, Asthma & Immunology