Objectives <p>We aim to explore the factors that influence the varying prognostic outcomes of itraconazole treatment in patients with onychomycosis. A prediction model for treatment failure of onychomycosis to itraconazole was developed and validated.</p> Methods <p>This study is a prospective, case-cohort study of observational cases. Patient information meeting the inclusion criteria will be grouped and compared based on their outcomes six months after treatment.</p> Results <p>A total of 437 patients were enrolled in our study, and the following factors were correlated with prognosis: patient ages, duration of illness, number of affected nails, fungal species, clinical classification, nail thickness, nail involvement, and bathing frequency. Based on these variables, we constructed a predictive model for treatment failure in onychomycosis. The accuracy and precision of the prediction model for itraconazole treatment failure in onychomycosis were 0.936 and 0.924.</p> Conclusion <p>First, this study validated that multiple factors are associated with the prognosis of itraconazole pulse therapy for onychomycosis, suggesting that early treatment of the disease and attention to bathing habits can lead to better therapeutic outcomes. Second, based on the contribution of representative variables to the predictive power of the binary multivariate logistic regression model, it can help clinicians better understand the characteristics of patients with onychomycosis who experience treatment failure. These characteristics mainly include a longer disease duration, non-dermatophyte infections, multiple nail involvement, and significant psychological impact. Based on these features, it provides a reliable basis for determining the optimal treatment course of itraconazole pulse therapy for onychomycosis and whether additional treatment is needed.</p>