The purpose of this study was to develop a risk prediction model for patients with nonobstructive CAD. Among stable chest pain patients, most cardiovascular (CV) events occur in those with nonobstructive coronary artery disease (CAD). Thus, developing tailored risk prediction approaches in this group of patients, including CV risk factors and CAD characteristics, is needed. In PROMISE (Prospective Multicenter Imaging Study for Evaluation of Chest Pain) computed tomographic angiography patients, a core laboratory assessed prevalence of CAD (nonobstructive 1% to 49% left main or 1% to 69% stenosis any coronary artery), degree of stenosis (minimal: 1% to 29%; mild: 30% to 49%; or moderate: 50% to 69%), high-risk plaque (HRP) features (positive remodeling, low-attenuation plaque, and napkin-ring sign), segment involvement score (SIS), and coronary artery calcium (CAC). The primary end point was an adjudicated composite of unstable angina pectoris, nonfatal myocardial infarction, and death. Cox regression analysis determined independent predictors in nonobstructive CAD. Of 2,890 patients (age 61.7 years, 46% women) with any CAD, 90.4% (n = 2,614) had nonobstructive CAD (mean age 61.6 yrs, 46% women, atherosclerotic cardiovascular disease [ASCVD] risk 16.2%). Composite events were independently predicted by ASCVD risk (hazard ratio [HR]: 1.03; p = 0.001), degree of stenosis (30% to 69%; HR: 1.91; p = 0.011), and presence of ≥2 HRP features (HR: 2.40; p = 0.008). Addition of ≥2 HRP features to: 1) ASCVD and CAC; 2) ASCVD and SIS; or 3) ASCVD and degree of stenosis resulted in a statistically significant improvement in model fit (p = 0.0036; p = 0.0176; and p = 0.0318; respectively). Patients with ASCVD ≥7.5%, any HRP, and mild/moderate stenosis had significantly higher event rates than those who did not meet those criteria (3.0% vs. 6.2%; p = 0.007). Advanced coronary plaque features have incremental value over total plaque burden for the discrimination of clinical events in low-risk stable chest pain patients with nonobstructive CAD. This may be a first step to improve prevention in this cohort with the highest absolute risk for CV events.