Adolescent Idiopathic Scoliosis (AIS) is a three-dimensional spinal deformity for which classification systems are fundamental to surgical planning and prediction of curve behavior. The Lenke classification system is the most widely adopted framework for AIS; however, increasing evidence suggests substantial heterogeneity within Lenke curve types and the presence of reverse curves that may not be fully captured by the classification. This retrospective study analyzed the distribution of AIS curve types and subclassifications over a 9-year period at a single tertiary institution. Consecutive AIS patients aged 10–18 years who underwent posterior spinal fusion between 2017 and 2025 were included. Erect posteroanterior and lateral whole-spine radiographs and supine side-bending radiographs were reviewed, curves were classified according to the Lenke system, including lumbar and sagittal modifiers. Additional subclassifications, including Miyanji 1AR/1AL and 2AR/2AL patterns and flexibility-based categorization for Lenke types 1 and 5 were applied. A total of 1008 patients were included and analyzed. The Lenke classification was applicable in 96.3% of cases (n = 971), with 3.7% (n = 37) unclassifiable. Among unclassifiable curves, Lenke 5–related patterns were most frequent (59.5%). Lenke 1 was the most prevalent curve type (44.9%), followed by Lenke 5 (22.9%). Lumbar modifier C predominated (48.5%), and normal thoracic sagittal alignment (N) was most common (79.3%). Subclassification revealed substantial heterogeneity within curve types: Lenke 1 curves showed an almost equal distribution of stiff and flexible patterns (50.2% vs 49.8%), whereas Lenke type 5 curves were predominantly flexible (67.1%). AR subtypes were more prevalent than AL subtypes in Lenke 1 (62.4% vs 37.6%) and Lenke 2 (67.6% vs 32.4%). Notably, several Lenke subtype combinations were absent, including 3A−/3A+, 3B−/3B+, 3C+, 4A−/4A+, and 4B−/4B+. In conclusion, while the Lenke classification system demonstrates high applicability in surgically treated AIS, clinically meaningful heterogeneity and a subset of reverse curves remain insufficiently characterized, underscoring the need for refined classification approaches to support more individualized surgical decision-making