OBJECTIVE: Our aim was to determine, using CT attenuation values, the chemical composition of 241 human renal stones placed in a jelly phantom and to analyze the influence of respiratory motion on the classification. MATERIALS AND METHODS: The stones were placed in a jelly simulating the X-ray attenuation of the kidneys. A dynamic platform was used to apply to the phantom free-breathing motion (sinusoidal motion in z-axis) and motion due to lack of maintenance of a breath-hold (5 mm x s(-1) in z-axis). Determination of the chemical composition was performed with mean CT attenuation values obtained at 80 and 120 kV and with dual-energy CT attenuation values. RESULTS: Two hundred forty-one human urinary stones were classified into six groups: uric acid, cystine, struvite, weddellite (calcium oxalate dihydrate), whewellite (calcium oxalate monohydrate), and brushite. With no motion, the use of dual energy enabled differentiation of all of the types of stones with statistically significant differences. Uric acid (-20 +/- 22 H), cystine (106 +/- 19 H), struvite (271 +/- 16 H), weddellite (323 +/- 5 H), brushite (415 +/- 30 H), and whewellite (510 +/- 17 H) were identified as distinct groups. Motion-induced mean CT attenuation values were significantly different from those obtained with no motion. With motion, dual-energy CT attenuation values did not allow differentiation of all stone types. CONCLUSION: Dual-energy CT attenuation values can be used to predict the chemical composition of stones in vitro. However, when slight motion is applied to renal stones during image acquisition, the values become significantly different from those obtained with no motion. Consequently, confusion arises in differentiating stone types. A perfect breath-hold has to be performed for in vivo use of attenuation value to discern stone type.