Although liver resection is the accepted treatment for patients with metastatic colorectal cancer to the liver, there remains some controversy as to the criteria for patient selection and its impact on the recurrent disease following the operation. One hundred and sixteen patients underwent liver resection for metastaic colorectal carcinoma over the last 23 years at our institution. The actual survival was 50.5% at 5 years, and 38.6% at 10 years, excluding the 30-day operative mortality rate of 1.7%. Positive hepatic lymph nodes, extrahepatic disease, number of tumors (4 or more), and bilobar distribution of the diseases were strongly associated with poor outcome. These prognostic factors were also documented as the major prognostic determinants by 21 consecutive articles we reviewed. Among 52 (45.6%) patients with hepatic recurrence, patients who were acceptable surgical risks and had no extrahepatic diseases or a limited number (one or two) of lung metastasis were candidates for the repeat resection. Consequently, 31 patients underwent repeated hepatic resection. Their survival rate was 53.9% at 3 years and 32.3% at 5 years, with no mortality. This outcome is markedly better than that of untreated patients with an estimated survival of 3 to 24 months. Thus, an aggressive approach to surgical treatment is feasible with low risk and the potentially curative strategy even for recurrent hepatic metastases, providing the opportunity to achieve 5 year survival.