Objectives <p>This study compared the silicone replica method with swept-source optical coherence tomography (SS-OCT) to analyze marginal adaptation and investigated the effect of the light incidence angle of SS-OCT on measurement precision.</p> Materials and methods <p>A typodont-prepared mandibular right first molar was scanned using an intraoral scanner (Trios 3). Fourteen crowns were fabricated from CAD/CAM resin blocks (Katana Avencia P) using a 5-axis milling machine (DWX-50). Marginal adaptation at the buccal, lingual, mesial, and distal points was assessed using the silicone replica method and SS-OCT at light incidence angles of 60°, 75°, and 90°. Statistical comparisons were performed using two-way analysis of variance (ANOVA) and t-tests with Bonferroni correction, and t-tests at a significance level of 0.05.</p> Results <p>At 60°, SS-OCT showed significantly larger marginal discrepancies than the silicone replica method at the buccal, lingual, and mesial points (<i>p</i> &lt; 0.05). At 75°, only the lingual point showed a significantly larger value than the silicone replica method (<i>p</i> &lt; 0.05). At 90°, no significant differences were observed between the SS-OCT and silicone replica methods (<i>p</i> &gt; 0.05). Marginal discrepancy values increased as the angle changed from 90° to 75° to 60°, with significant differences between 60° and 75° and between 60° and 90° at the buccal and lingual points (<i>p</i> &lt; 0.05).</p> Conclusions <p>SS-OCT is a viable alternative to the silicone replica method for assessing marginal adaptation at an incidence angle of 90 °.</p> Clinical relevance <p>SS-OCT, a non-invasive method, has the potential to be applied clinically for evaluating marginal fit in indirect restorations in vivo.</p>