Background: Tumor grade is a predictor of breast cancer recurrence, independent of gene expression assays such as Oncotype Dx (ODx). However, grade suffers from significant intra- and inter-observer variability, limiting its clinical utility. Advances in whole slide imaging facilitate computerized analysis of H&E slides. We have developed a prognostic assay called Image-based Risk Score (IbRiS) which extracts and leverages morphometric signatures from H&E slides to classify patients into low- vs. high-risk groups for recurrence. Methods: IbRiS was evaluated on 378 ER+ patients treated with chemo-hormonal therapy obtained from ECOG-ACRIN study E2197; patients had 0-3 positive axillary nodes and H&E slides of acceptable quality; 60 patients suffered recurrence. A subset of 124 patients (27 recurrences) were HER2-. Another (enriched) subset (116 patients with 27 recurrences) had accompanying ODx scores. To establish IbRiS low- and high-risk groups, we determined (over the 378 patients) a cutoff such that the low-risk group had a 10-year recurrence rate of ~10%. We evaluated IbRiS risk groups using Cox proportional hazard analysis, controlling for node status, age, and tumor size. We also compared the recurrence rates in the IbRiS risk groups to those of ODx and a combined assay (IbRiS + ODx). Results: IbRiS risk groups predicted recurrence with a hazard ratio of 2.41 (n = 378, 95% CI = 1.21-4.79, p = 0.01) and 2.52 (n = 124, 95% CI = 0.85-7.46, p = 0.09) for the HER2- subset. Over the 378 patients, IbRiS classified as low-risk 35.3% of the patients who did not suffer recurrence. Table 1 compares the low-/high- risk stratifications of IbRiS, ODx, and the combined assay. Conclusions: IbRiS assay is a significant predictor of breast cancer recurrence and contributes prognostic information independent of ODx. The combined genomic and morphometric assay outperforms either individual assay. Also, while ODx was reported to lose prognostic ability when restricted to HER2- patients in E2197, IbRiS remains a predictor of recurrence.