The purpose of this study was to investigate the feasibility and safety of lung radiofrequency (RF) ablation by using low-perfusion-rate, expandable, multitined electrodes in an in vivo animal model. Ten New Zealand White rabbits underwent RF ablation using low-perfusion-rate, expandable, multitined electrodes (Starburst Talon; RITA Medical Systems, Mountain View, CA) and a 200-W RF generator. The electrode was positioned under fluoroscopy guidance and a single percutaneous RF ablation was performed. Saline perfusate was doped with nonionic iodinated contrast agent to render it visible on computed tomography (CT). The pump infused the saline doped with contrast agent into the lateral tines at a rate of 0.1 ml/min. The planned ablation was of 3 min, with the hooks deployed to 2 cm at a target temperature of 105 degrees C. An immediate posttreatment CT scan documented the distribution of the doped saline and the presence of immediate complications. The animals were monitored for delayed complications and sacrificed within 72 h (n = 4), 2 weeks (n = 3), or 4 weeks (n = 3). Assessment of ablation zone and adjacent structures was done at autopsy. Major complications consisted of pneumothorax requiring drainage (n = 2) and skin burn (n = 1). Immediately after the procedure the area of ablation was depicted at CT as a round, well-demarcated area, homogeneously opacified by iodinated contrast medium (mean size, 2.3 +/- 0.8 cm). The presence of a sharply demarcated area of coagulation necrosis (mean size, 2.1 +/- 0.4 cm) without severe damage to adjacent structures was confirmed at autopsy. In one case, euthanized at 4 weeks, in whom pneumothorax and pleural effusion were depicted, pleural fibrinous adhesions were demonstrated at autopsy. In conclusion, lung RF ablation performed in an in vivo animal model using low-perfusion-rate, expandable, multitined electrodes is feasible and safe. No severe damage to adjacent structures was demonstrated.