PURPOSE: To compare histopathologically the completeness of radiofrequency (RF) ablation to treat hepatocellular carcinoma (HCC) with monopolar or multipolar technique. MATERIALS AND METHODS: Thirty-five consecutive patients (mean age, 59 y) with cirrhosis and HCC (n = 59) within Milan criteria received RF ablation and subsequently underwent liver transplantation (LT) for tumor progression or liver failure. Data were extracted retrospectively from a prospective database. Thirty nodules were treated with a monopolar device with internally cooled (n = 17) or perfused (n = 13) electrodes, and 29 were treated with a multipolar technique with internally cooled electrodes based on the "no-touch" concept. This consisted of inserting two or three straight electrodes around the nodule to avoid intratumor puncture to the greatest extent possible. Effectiveness of the three devices was compared by histopathologic examination of explants. Fisher exact and χ(2) tests and multivariate logistic regression analysis were performed. RESULTS: Mean sizes of nodules ablated (25, 22, and 21.6 mm) and median times from ablation to LT (11, 7.5, and 8.4 months) for patients treated with the monopolar internally cooled electrode device (MoICD), monopolar perfused electrode device (MoPED), and multipolar internally cooled electrode device (MuICD), respectively, were similar (P = .8 and P = .9, respectively). Pathologic examination showed complete necrosis for eight of 17 and six of 13 nodules treated with the MoICD and MoPED, respectively, versus 26 of 29 treated with the MuICD (P = .0019). In multivariate analysis, RF technique remained the predictive factor for complete necrosis (P = .005). CONCLUSIONS: Ablation of small HCCs with multipolar RF ablation based on the no-touch concept improves the rate of complete necrosis during pathologic examination compared with monopolar techniques.