Background <p>Evidence on nonoperative treatment for Legg-Calvé-Perthes disease remains limited by heterogeneity in treatment exposure, including variable combinations of observation, physiotherapy, bracing, casting, activity restriction, and weight-bearing restriction. We evaluated radiographic outcomes and risk factors after a protocol-defined prolonged cast-based abduction containment regimen, thereby establishing benchmark outcome estimates for this specific nonoperative pathway.</p> Methods <p>This single-center retrospective cohort study included children with Legg-Calvé-Perthes disease treated with the institutional protocol between 2009 and 2022 and followed radiographically to healing. The protocol consisted of continuous cast-based abduction containment and strict non-weight-bearing until radiographic healing. The primary outcome was an adverse radiographic result at healing, defined as modified Stulberg grade III to V. Hip-level associations were evaluated using generalized estimating equations with an independence working correlation structure and patient-clustered robust standard errors to account for correlation between bilateral hips.</p> Results <p>A total of 398 children with 427 hips were included; 110 hips (25.8%) had an adverse radiographic outcome. Adverse outcome rates increased across age groups: 15.8% in children younger than 6 years, 24.1% in children aged 6 to &lt; 8 years, and 39.8% in children aged 8 years or older. In the primary multivariable model, compared with children younger than 6 years, the odds of adverse outcome were higher in children aged 6 to &lt; 8 years (OR 2.16, 95% CI 1.13 to 4.15; <i>p</i> = 0.020) and in those aged 8 years or older (OR 5.69, 95% CI 2.77 to 11.71; <i>p</i> &lt; 0.001). Compared with Herring A/B hips, Herring B/C hips (OR 2.19, 95% CI 1.07 to 4.49; <i>p</i> = 0.032) and Herring C hips (OR 10.28, 95% CI 5.50 to 19.22; <i>p</i> &lt; 0.001) had higher odds of adverse outcome. Sensitivity analyses showed the same overall pattern.</p> Conclusions <p>Within this defined nonoperative pathway, older age at diagnosis and greater Herring lateral pillar involvement were associated with a higher risk of adverse radiographic outcome at healing. These protocol-specific estimates provide a reference for prognostic counselling within this treatment pathway and a benchmark for future comparative studies of nonoperative strategies.</p> Trial registration <p>Chinese Clinical Trial Registry, ChiCTR2600120879, 20 March 2026, retrospectively registered.</p>