Purpose <p>To evaluate the association between the newly developed region of interest (ROI)-modified Mayo Adhesive Probability (MAP) score, in which stranding was re-evaluated by computed tomography (CT) number, for predicting operation time in robot-assisted partial nephrectomy (RAPN).</p> Methods <p>The study participants were 119 patients who underwent transperitoneal RAPN. With regard to stranding, ROIs were evaluated, and the mean CT numbers were assigned a score ranging from 0 to 3. Clinical variables were evaluated in a multivariate logistic regression analysis in relation to prolonged operation time.</p> Results <p>The percentage of patients with score ≥ 3 by MAP score alone was significantly higher than those of patients with score ≥ 3 by ROI-modified MAP score alone (26.8% vs. 13.4%, <i>p</i> &lt; 0.001). Multivariate analysis revealed no independent association with the MAP score. On the other hand, for ROI-modified MAP score, score ≥ 3 was an independent factor for prolonged operation time (OR = 4.28, <i>p</i> = 0.0032) along with body mass index (BMI) ≥ 22 (OR = 4.46, <i>p</i> = 0.01), R.E.N.A.L. nephrometry score ≥ 7 (OR = 4.12, <i>p</i> = 0.0047), posterior tumor location (OR = 2.85, <i>p</i> = 0.036), and clinical T stage ≥ 1b (OR = 6.19, <i>p</i> = 0.0044). Regarding the predictive performance, the accuracy of the ROI-modified MAP score was significantly higher than the MAP score (area under the curve [AUC] value: 0.652 vs. 0.721, <i>p</i> = 0.034).</p> Conclusion <p>The ROI-modified MAP score was a more relevant factor regarding operation time, suggesting that it might be a better preoperative predictor.</p>