BACKGROUND An analysis of perioperative factors that were independently associated with survival up to 1000 days after scheduled colorectal adenocarcinoma resections in 314 patients. METHODS The association of 16 perioperative variables with postoperative survival and critical care unit (CCU) admission after scheduled resections for colorectal adenocarcinoma between September 2005 and March 2009 was analysed using multivariable Cox regression analyses and Fisher's exact tests. RESULTS We followed survivors for a mean of 983 days (range 696-1000 days). Average annual postoperative mortality was 8.5%, 14% after surgery performed in 2005, and 3% after surgery in 2009. Risk of mortality was independently associated with five variables after stratifying for date of surgery: attending a preoperative high-risk clinic [hazard ratio (HR) 0.42, P=0.006], worse World Health Organization performance status (HR 2.1, P=0.001), BMI (HR 0.92, P=0.009), higher nodal stage (HR 2.6, P<0.0001), and unplanned critical care admission (HR 7.2, P<0.0001). Patients who attended the preoperative clinic (207) were older, with worse renal function and ASA grade, than those who did not (107). Planned postoperative critical care admission was more common in patients who attended the high-risk clinic (24% vs 12%, P=0.01) and may have partly accounted for the observed mortality difference. CONCLUSIONS Immediate perioperative care may have prolonged effects on postoperative survival. Specialized preoperative assessment clinics may reduce mortality after colorectal surgery.