BACKGROUND: Although the benefits of helmet therapy for positional plagiocephaly are strongly correlated with age, the effective period remains controversial. However, most physicians agree that effective results can be obtained in patients within the age of 6 months. Owing to the characteristics of positional plagiocephaly in Koreans, many Korean patients have delayed diagnosis, and because this results in delayed onset of the helmet therapy, the outcomes remain largely underevaluated. In the management of late-diagnosed positional plagiocephaly, we aimed to determine the factors affecting the effective application of helmet therapy. METHODS: We recruited 39 consecutive patients with positional plagiocephaly who received helmet therapy and completed the treatment between December 2008 and June 2016. The ages at initiation and completion of treatment, duration of daily use, initial and final absolute diagonal differences, cephalic index, and cranial vault asymmetry index (CVAI) were analysed using data retrospectively collected from the patients' medical records. RESULTS: We identified 12 patients with late-diagnosed positional plagiocephaly, of whom 83.33% were effectively treated. The effective change in CVAI (%) was affected by age at treatment initiation (P = 0.001), initial absolute diagonal distance differences (P < 0.001), and initial CVAI (P < 0.001). Up to 9 months, a gradual change of at least 1% CVAI was attained. Treatment initiation at ages < 5.5 months was beneficial. Even at a later age, patients with an initial absolute diagonal distance difference of > 13.50 mm and initial CVAI of > 11.03% could receive effective helmet therapy. CONCLUSION: The efficacy of helmet therapy in late-diagnosed patients can be predicted on the basis of not only age at treatment initiation, but also initial absolute diagonal distance differences and initial CVAI. We anticipate that even patients with late-diagnosed positional plagiocephaly can expect better helmet therapy outcomes.