Objectives: To identify potential underlying mechanisms of early and (very) late scaffold thrombosis (ScT) by optical coherence tomography (OCT), in a frame-by-frame analysis. Background: The absorb scaffold is associated with an increased risk of ScT compared with metallic stents. Several potential causes of bioresorbable ScT have been identified, however the precise etiology still remains unclear. Methods: Between February 2013 and February 2016, 13 patients presenting with definite ScT underwent OCT imaging. After guidewire passage or balloon inflations, OCT images were acquired. Pullbacks were assessed offline at each 1 mm longitudinal interval within the treated segment and the 5 mm segments adjacent to both edges. Primary cause of ScT was assessed by reviewing medical records, baseline angiographic films, and OCT pullback and angiographic films at time of ScT. Results: 13 patients, with 14 thrombotic lesions presented either with early ScT (i.e., ≤30 days) or very (late) (i.e., >30 days). Analysis demonstrated a significantly smaller in-scaffold maximal lumen diameter in the early cases (2.75 ± 0.85 mm vs. 3.00 ± 0.46 mm; P = 0.033) and a nonsignificant smaller minimal scaffold diameter (2.44 ± 0.62 mm vs. 2.58 ± 0.37 mm P = 0.097). Per-strut analysis demonstrated significantly more malapposed scaffold struts in (very) late cases (6% versus 0.6%, P < 0.001). Assessment of the predominate cause showed underexpansion as the dominant factor in the early cases, while malapposition was predominantly seen in the (very)late cases. Conclusions: OCT performed in patients presenting with Absorb ScT demonstrated that malapposition of scaffold struts was more prominent in patients presenting with (very) late ScT, while underexpansion was more frequent in the early cases.