To evaluate factors that affect the size of the thermal lesion created from radiofrequency ablation (RFA) with an impedance-based system. Thirty-two nonresectable liver tumors, including hepatocellular carcinomas (HCC) (N = 20; 63%) or metastatic tumors (N = 12; 37%) were treated in 29 patients with an impedance based RFA system. Tumor diameter was measured at pre-procedure cross sectional imaging. Follow-up imaging was obtained at 1 month and every 3 months afterward to determine the short axis of the resultant thermal lesion. Success at durable tumor destruction was tracked as well as size of the resulting thermal lesion and the resulting thermal lesion/electrode ratio. Time to impedance roll-off was also correlated with treatment success. Twenty-eight (88%) of the 32 tumors were durably ablated in a single session. Two patients had residual disease and two had local recurrence. All residual or recurrent lesions were in patients with cirrhosis. Three of these patients were retreated resulting in a durable result in 97% of lesions with a mean/median follow up 443/503 days. Mean tumor diameter and resulting thermal lesion size was similar for HCC and metastatic lesions. A significant difference (P = .02) was observed when comparing the thermal lesion/electrode ratios of HCC (1.01 ± 0.36) with metastases (1.26 ± 0.22). Nine of 20 HCC ablations (45%) resulted in a ratio > 1.0, whereas 11 of 12 metastatic ablations (92%) attained a ratio > 1.0 (P = .01). Thermal lesion size did not correlate with time to impedance roll-off for the entire group (r =-0.22; P = .90) or by tumor type (HCC r = 0.10; P = .68; metastases r =-0.36; P = .23). Thermal lesion size and the ability to obtain a margin of normal tissue are significantly affected by the presence of cirrhosis when using an impedance-based system. Given this limitation, the majority of small hepatic neoplasms can still be successfully treated with RFA.