OBJECTIVE:: This study hypothesized that tumor size, number of tumors, surgical approach, and tumor histology significantly affected microwave ablation (MWA) success and recurrence-free survival. BACKGROUND:: Although many hepatobiliary centers have adopted MWA, the factors that influence local control are not well described. METHODS:: Consecutive patients with hepatic malignancy treated by MWA were included from 4 high-volume institutions (2003-2011) and grouped by histology: hepatocellular carcinoma (HCC), colorectal liver metastases, neuroendocrine liver metastases, and other cancers. Independent significance of outcome variables was established with logistic regression and Cox proportional hazards models. RESULTS:: Four hundred fifty patients were treated with 473 procedures (139 HCC, 198 colorectal liver metastases, 61 neuroendocrine liver metastases, and 75 other) for a total of 875 tumors. Median follow-up was 18 months. Concurrent hepatectomy was performed in 178 patients (38%), and when performed was associated with greater morbidity. Complete ablation was confirmed for 839 of 865 tumors (97.0%) on follow-up cross-sectional imaging (10 were unevaluable). A surgical approach (open, laparoscopic, or percutaneous) had no significant impact on complication rates, recurrence, or survival. The local recurrence rate was 6.0% overall and was highest for HCC (10.1%, P = 0.045) and percutaneously treated lesions (14.1%, P = 0.014). In adjusted models, tumor size 3 cm or more predicted poorer recurrence-free survival (hazard ratio: 1.60, 95% CI: 1.02-2.50, P = 0.039). CONCLUSIONS:: In this large data set, patients with 3 cm or more tumors showed a propensity for early recurrence, regardless of histology. Higher rates of local recurrence were noted in HCC patients, which may reflect underlying liver disease. There were no significant differences in morbidity or survival based on the surgical approach; however, local recurrence rates were highest for percutaneously ablated tumors.