<p>Hay fever remains a global health concern, with cedar pollen being Japan’s primary seasonal allergen affecting one-third of the population. The COVID-19 pandemic led to behavioral changes, including mask use and social distancing, with reports of reduced allergic symptoms. To evaluate changes in hay fever symptoms and related preventive behaviors before and during the COVID-19 pandemic using self-reported data from a smartphone application. This digital cross-sectional study used the AllerSearch application to collect self-reported data from Japanese residents with hay fever symptoms who provided informed consent. The population was divided into pre-pandemic and mid-pandemic cohorts. Outcomes included total symptoms score, nasal symptoms score, non-nasal symptoms score, quality of life&#xa0;(QoL), and stress levels. Multivariable analysis, machine learning, and exploratory causal analysis using Linear Non-Gaussian Acyclic Model (LiNGAM) were conducted. Hay fever symptoms worsened during the mid-pandemic, with increased mask usage but decreased use of goggles and oral medications. The mid-pandemic period, higher stress levels, impaired QoL, younger age, higher temperatures, and elevated pollen counts were significant symptom severity predictors. LiNGAM identified pollen count, particulate matter&#xa0;2.5, and female sex as directional associations, while older age was associated with reduced symptoms. The COVID-19 pandemic influenced hay fever symptoms and preventive behaviors, informing future strategies for managing seasonal allergies.</p>