Background: Allergic rhinitis (AR) is characterized by distinct clinical heterogeneity and allergic sensitization patterns. We aimed to quantify rhinitis symptoms in patients with self-reported allergic rhinitis according to the potential sensitization patterns for relevant allergens in China. Methods: We used latent class analysis (LCA; a subset of structural equation modeling) to independently cluster patients into different patterns of atopic sensitization in an unsupervised manner, based on specific immunoglobulin E tests. AR symptom severity was assessed by the visual analogue scale. We evaluated the association between the severity of AR and the allergen sensitization patterns. Results: LCA revealed four phenotypes of atopic sensitization among 967 patients with self-report AR. We labeled latent classes as: Class 1, weed pollens and indoor sensitization (n = 74 [7.7%]); Class 2, weed pollen with low indoor sensitization (n = 275 [28.4%]); Class 3, low or no sensitization (n = 350 [36.2%]); and Class 4, house dust mite-dominated sensitization (n = 268 [27.7%]). AR was more severe in Class 2 compared to the other 3 classes, indicating that upper respiratory symptoms are more severe among patients with isolated seasonal rhinitis. Conclusion: We have identified four sensitization patterns in patients with self-reported AR, which were associated with different clinical symptoms and comorbidities.