PURPOSE: The purpose of this work was to determine the reproducibility of coronary total calcium score (TCS) with dual-slice helical CT and compare three acquisition protocols. METHOD: Fifty patients (59 +/- 10 years old) underwent dual-slice helical CT (collimation = 2 x 2.5 mm) and coronary angiography. Two successive scans were performed, resulting in three sets of images: pitch = 1, 360 degrees linear interpolation (LI) (A360); pitch = 1, 180 degrees LI (A180); and pitch = 1.5, 180 degrees LI (B180). TCS values, calculated using a volumetric method with a threshold of 90 HU, were compared, and the interscan variation was determined. Diagnostic performances were compared with receiver operating characteristic curves. RESULTS: Protocol A360 provided significantly lower TCS than protocols A180 and B180 (p < 0.0001). No statistical difference was seen between A180 and B180, which provided the lowest interscan variation (40 +/- 58%). However, no significant clinical impact of the observed interscan variations was found. CONCLUSION: Reproducibility of TCS with dual-slice helical CT is improved by the 180 LI algorithm. However, dual-slice helical CT is not sufficiently reproducible to allow serial quantification of TCS over time.