Abstract:AIM: To compare in vivo confocal microscopy (IVCM) and anterior segment optical coherence tomography (AS-OCT) in evaluating demarcation line depth (DLD) after corneal collagen cross-linking (CXL) performed under different ultraviolet (UV) energy levels in keratoconus. METHODS: This was a retrospective case series. This study included 216 participants (328 eyes) who underwent CXL with either high (7.2 J/cm²) or low (1.8 J/cm²) UV energy. DLD was assessed by IVCM and AS-OCT at 3, 6, and ≥12mo postoperatively. The high-energy group consisted of participants with progressive keratoconus who received CXL treatment, while the low-energy group included participants with refractive errors who underwent femtosecond laser-assisted in situ keratomileusis (FS-LASIK) Xtra. RESULTS: The high-energy group included 111 participants (144 eyes, mean age 23.39±7.30y, 68 males), and low-energy group included 105 participants (184 eyes, mean age 22.95±5.84y, 53 males). At 3 and 6mo, as well as at the final follow-up, visibility of the demarcation line was consistently superior with IVCM compared with AS-OCT (all P<0.001). At 3mo, the mean DLD measured by IVCM was 382.2±91.12 μm in the high-energy group and 192.53±58.76 μm in the low-energy group, while AS-OCT recorded 354.74±78.82 μm and 196.51±55.57 μm, respectively. Both modalities demonstrated significantly deeper demarcation line with higher UV energy (P<0.05), and no significant difference was observed between IVCM and AS-OCT measurements within each energy group (P>0.05). Furthermore, a significant monotonic temporal trend toward shallower IVCM-detected DLD was observed in both energy groups. CONCLUSION: Compared with AS-OCT, IVCM provides clearer and more reliable visualization of the corneal stromal demarcation line after CXL. While both modalities yield comparable depth measurements, IVCM maintains higher visibility across different energy levels and during follow-up at various time points, making it a preferred modality for postoperative assessment in clinical practice.