Objective: We aimed to investigate diagnostic accuracy of the STONE score and the modified STONE score by external validation in a large-scale cohort. Methods: We retrospectively reviewed the medical records of patients with suspected urinary stones who underwent computed tomography in the emergency center of a single tertiary hospital from 2014 to 2015. We compared and analyzed our cohort and two original studies with each other. Patients were categorized into three risk groups (low, moderate, and high) according to both scoring systems. The prevalence of urinary stones and alternative diagnoses was evaluated in each group. We calculated sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) with our data for each score performance. Results: Eight hundred fifty-six patients out of a total of 1,013 were diagnosed with urinary stones. The prevalence of urinary stones in the high-risk group of the STONE and the modified STONE scores were 91.1% and 91.2%, respectively. The areas under the receiver operating characteristic curve of both scores were 0.71 and 0.73, respectively. The optimal cutoff value for predicting urinary stones using the STONE score was 9, with 87.9% sensitivity, 45.9% specificity, 89.8% PPV, and 40.9% NPV. In the case of the modified STONE score, the optimal cutoff value was 10, which was the same as that of the original study with 85.9% sensitivity, 54.8% specificity, 91.2% PPV, and 41.6% NPV. Conclusion The STONE and the modified STONE scores showed good diagnostic accuracy in high-risk groups of patients with suspected urinary stones. These scoring systems would be helpful to physicians in their diagnoses and in reducing radiation exposure.