A 63-year-old man with hepatocellular carcinoma consequent to chronic viral hepatitis C presented with severe dermatitis on the lower right side of the back after transcatheter arterial chemoembolization (TACE) via the 10th intercostal artery, because his hepatic artery had already collapsed due to repeated usage for TACE. The regional skin showed redness and hardness with pustules. Histologically, there was epidermal and appendage necrosis, as well as exocytosis of red blood cells. With a diagnosis of skin injury due to leakage of lipiodol ultrafluid, mitomycin, and epirubicin, administered via the 10th intercostal artery, onto the skin tissue, topical application of 0.06% fluocinonide-containing cream was prescribed. After several weeks of conservative local treatment, the leakage skin injury improved significantly, leaving pigmentation, hardness, and a small necrotic mass, as reported elsewhere (Honda T, Matsushima S, Fujii S, et al. A case of skin injury following transcatheter arterial chemotherapy through intercostal artery for hepatocellular carcinoma. Skin Res 2003; 2: 18-22). Subsequently, the patient again consulted the Dermatology Department with further dermatitis in an almost identical skin region on the right side of the abdomen (irregularly spreading erythematous and edematous eruptions with itching; Fig. 1). As he had undergone an ultrasonic examination 2 days earlier, allergic contact dermatitis from the ultrasonic gel was suspected. The contact dermatitis was treated with a topical corticosteroid hormone-containing ointment. Patch testing was performed with Ultra Phonic Conductivity Gel (Pharmaceutical Innovations Inc., Newark, NJ, USA), with which the patient had undergone a series of ultrasonic examinations, and Sono Jelly (Toshiba Medical Supply Co., Ltd., Tokyo, Japan) as a reference, as well as white petroleum as a negative control. A positive result was obtained for Ultra Phonic Conductivity Gel, whereas Sono Jelly was negative (Fig. 2a). Pharmaceutical Innovations Inc. kindly supplied the ingredients of the gel: propylene glycol (PG), preservative in PG, color in PG, thickener 1, and thickener 2. The company gave no further details about the preservative, color, and thickeners. Patch testing was performed using these five materials, resulting in a positive reaction for PG, preservative in PG, and color in PG. Thickeners 1 and 2 and lipiodol ultrafluid were negative (Fig. 2b). On the assumption that the causative chemical was PG, commercially obtained PG, free of preservative and color, was then patch tested (original, 10% aqueous, 1% aqueous, and 0.1% aqueous solutions), resulting in original strongly positive, 10% positive, and 1% and 0.1% weakly positive (Fig. 2c). The medical records showed that the patient had received 16 ultrasonic examinations with the same ultrasonic gel before the leakage skin injury. The 17th examination was performed 2 days after leakage dermatitis, and the 18th 3 months after the injury, when contact dermatitis occurred. The 19th examination was performed using Sono Jelly, which contains no PG, and no skin problems were observed.