The usefulness of interventional radiology (IVR) in clinical practice is well known. However, patient dose in IVR has recently been increased as a result of the prolongation of fluoroscopic time and the increased number of radiographies. We studied a simple method of calculating skin surface dose in patients who underwent transcatheter arterial embolization (TAE) for the treatment of hepatocellular carcinoma by obtaining the value of a dose area product meter attached to the digital subtraction angiography system. In 20 subjects (15 men and 5 women, aged an average of 68.2+/-7.3 years, respectively) who underwent TAE, exposure conditions (tube voltage, tube current, time, and size of image intensifier) in a time series and last value indicated on the dose area product meter were recorded. A dosimetric phantom was placed at a position the same as that of the patient for TAE, the surface dose (SD) of the phantom was measured under various exposure conditions, and SD per unit mAs (SD/mAs) was obtained. Then the skin surface dose in each subject was estimated from the values of the exposure condition and SD/mAs. A high correlation was observed between the last value (x) on the dose area product meter and the estimated skin surface dose (y) (r=0.933), and the following regression equation was derived: y=0.005x-0.589. The skin surface dose calculated using the regression equation was compared with that obtained by the method recommended by the Japan Association on Radiological Protection in Medicine (JARPM), considering the value estimated from the value of exposure conditions with SD/mAs as the gold standard. The results indicated that the error in the method using the regression equation was significantly lower than that of the JARPM method (18.3+/-14.0% and 75.5+/-66.0%, respectively, p<0.01). In conclusion, the skin surface dose in TAE could be monitored with high precision using the value of the dose area product meter by obtaining the regression formula between the value of the dose area product meter and the skin surface dose estimated with the phantom values.