There is a pressing need for effective therapies to treat uveal melanoma. Agents that inhibit the c-MET pathway have shown promise in multiple malignancies that overexpress c-MET. Herein, we assess c-MET expression in both primary uveal melanoma and liver metastases of uveal melanoma and evaluate the association of c-MET expression with clinical and pathologic variables. We have retrospectively identified tumor samples from primary and liver metastases of uveal melanoma from 1 January 1990 to 1 January 2012. We utilized immunohistochemistry to assess c-MET expression, and two pathologists quantified c-MET expression using an H-score (product of the intensity of staining and percentage of positive cells). The Mann-Whitney U-test, Pearson's correlation, and Cox model were used as appropriate. Thirty-nine of 40 (98%) primary tumors and nine of 10 (90%) metastatic liver lesions expressed c-MET (H-score range 0-300). There was a strong association between the percentage of positive cells and the intensity of c-MET expression (P=0.007). We found no association between c-MET H-score and clinicopathologic variables such as age, sex, or stage. c-MET expression was significantly higher in metastatic compared with primary tumors (median H-score 190 vs. 30, P=0.022). c-MET is expressed in the vast majority of primary and liver metastases of uveal melanomas; however, c-MET expression did not associate with pathologic features in our cohort. Metastatic lesions have higher expression of c-MET expression than primary tumors. Clinical trials involving c-MET inhibitors deserve further study in patients with uveal melanoma in both the adjuvant and metastatic setting.