Introduction: Surgery of liver metastases has clear benefit in increases survival of patients with colorectal cancer. If resection must be clone on anatomical basis or only wedge resection Is still in discussion. The objective of this work is to analyze and to compare the anatomical liver resections (RA) versus the not-anatomical ones (RNA) in the treatment of liver metastases from colorectal cancer. Material and Method: Retrospective revision of all patients operated by colorectal liver metastases from year 2000 to the 2007. RA was considered according to the principles of Couinaud and that included one or more liver segments. The sub-segmental or wedge resections were denominated RNA. Results: Of a total of 164 resections 34, 7% (57) were by colorectal metastases. 61% were PA. 23% RNA and 16% combined. There was 19% of extended resection (5 or more segments). The global morbidity was 35% and mortality of 0%. When comparing the different variables, the operating time, blood transfusions, prolongation of the prothrombin time and UCI stay, were significantly greater In the RA with respect to the RNA. On the other hand, there was greater amount of R0 Surgeries in RA (p < 0.05). With respect to the bad prognosis criteria, in average the patients with RA had more criteria versus RNA patients (p < 0.05). The actuarial survival was 50% and 29% for RA and RNA respectively (p = ns). The free disease survival was 20% and 19% for RA and RNA respectively (p = ns). Discussion: The anatomical surgery of the liver allows removing a larger number and size of lesions with similar morbidity and mortality. Apparently there would be a benefit in survival In patients with RA but the difference is not significant in this study.