Abstract: Objective: We investigated the clinical value of subtraction coronary computed tomography angiography (sub-CCTA) in double breath-holding mode for assessing in-stent restenosis. Materials and methods: Records were identified for stent implantation patients who underwent coronary CTA between October 2022 and May 2023. Images of stented segments before and after subtraction were evaluated via a double-blinded protocol by two physicians with more than 10 years of cardiovascular image reading experience. The percentage of assessable stented segments before and after subtraction and the physician's diagnostic confidence were calculated and compared with invasive coronary angiography (ICA). Diagnostic accuracy was measured by area under the receiver operating characteristic curve, with ≥50% stenosis as the threshold, and the clinical value of sub-CCTA for assessing in-stent restenosis was analyzed using multiple statistical methods. Results: A total of 182 stented segments in 83 patients were analyzed. The number of non-assessable segments in sub-CCTA compared to con-CCTA decreased from 47 (29.2%) to 13 (8.1%) (p < 0.001), and physician diagnostic confidence increased from 2.55±1.1 to 3.86±0.7 (p=0.001 < 0.01). Using ICA results as the gold standard, 54 stent segments from 24 patients were included in a quasi-segmental accuracy study. Sub-CCTA increased diagnostic accuracy from 62.6% to 81.7% (p=0.03 < 0.05) and decreased the false-positive rate from 35.2% to 5.8%. Conclusions: Sub-CCTA significantly improved the accuracy of diagnosing in-stent restenosis and the interpretability of stent segment images, enhanced the diagnostic confidence of interpreting physicians, and reduced unnecessary invasive tests for patients.