Having three liver resections for colorectal metastases has long been considered to be associated with a high risk of postoperative complications. The present study was designed to assess the feasibility and survival benefits of three liver resections. Between 2004 and 2011, data for 273 consecutive patients with colorectal metastases were analyzed. The patient characteristics, tumor status, operation-related variables, degree of liver steatosis, and short- and long-term outcomes were compared according to the number of liver resections. The history of preoperative chemotherapy was higher for patients who had had three liver resections as compared with other resections: i.e., one resection 41.0 %, versus two resections 56.8 %, versus three resections 81.8 %; p = 0.04. Patients receiving three liver resections had a high rate of liver steatosis (17.9 vs. 32.4 vs. 59.1 %; p = 0.03). The median operative time for three resections was significantly longer than for the other resections (359 min [range: 115-579 min] vs. 395 min [range: 178-740 min], vs. 482 min [range: 195-616 min]; p = 0.04). However, the complication rate and the postoperative hospital stay did not differ among the three groups. The 1-, 3- and 5-year survival rates did not differ significantly among the three groups (83.3, 57.5, and 44.6 % for one resection vs. 92.3, 52.1, and 35.7 % for two resections vs. 93.3, 49.0, and 34.1 % for three resections). Patients who had <5 tumors at a third liver resection and a recurrence interval of ≥ 500 days from the second resection were good candidates for three resections. Undergoing three resections of colorectal metastasis is feasible and provides a similar survival benefit as one or two resections, without increasing morbidity or mortality.