Résumé : OBJECTIVE. The objective of our study was to evaluate in vivo urinary calculus characterization with third-generation dual-source dual-energy CT (DECT) at reduced versus standard radiation dose. SUBJECTS AND METHODS. One hundred ffty-three patients requiring unenhanced CT for suspected or known urolithiasis were prospectively included in our study. They underwent two acquisitions at reduced-dose CT (90 kV and 50 mAsref; Sn150 kV and 31 mAsref, where Sn denotes the interposition of a tin flter in the high-energy beam) and standard-dose CT (90 kV and 50 mAsref; Sn150 kV and 94 mAsref). One radiologist interpreted the reduceddose examinations before the standard-dose examinations during the same session. Among 103 patients (23 women, 80 men; mean age ± SD, 50 ± 15 years; age range, 18-82 years) with urolithiasis, dedicated DECT software measured the maximal diameter and CT numbers, calculated the DECT number ratio, and labeled with a color code each calculus visualized by the radiologist as uric acid (UA) or non-UA. Volume CT dose index (CTDIvol) and dose-length product (DLP) were recorded. RESULTS. The radiologist visualized 279 calculi on standard-dose CT and 262 on reduced-dose CT; 17 calculi were missed on reduced-dose CT, all of which were = 3 mm. Among the 262 calculi visualized at both doses, the CT number ratio was obtained with the software for 227 calculi and was not different between the doses (p = 0.093). Among these 262 calculi, 197 were labeled at both doses; 194 of the 197 labeled calculi were labeled with the same color code. Among the 65 remaining calculi, 48 and 61 (all = 5 mm) were not labeled at standard-dose and reduced-dose CT (p = 0.005), respectively. At reduced-dose CT, the mean CTDI vol was 2.67 mGy and the mean DLP was 102.2 mGy × cm. CONCLUSION. With third-generation dual-source DECT, a larger proportion of calculi = 5 mm are not characterized as UA or non-UA at a reduced dose.