Introduction: Most admissions to Intensive Care (ICU) are unplanned and associated with acute illness, however, co-morbidities also have an impact on survival. Patients with colorectal cancer represent the largest group of cancer patients admitted to ICU. It is unknown which factors are important at determining survival, and whether it is the features associated with the underlying malignancy or the acute illness that has the largest impact upon prognosis. Methods: We used routinely collected data from West of Scotland ICUs linked to Scottish Cancer registry data to identify patients (aged 16 and over) who had been admitted to ICU between 1st January 2000 and 31st December 2011 and who had a diagnosis of colorectal cancer (ICD10 coding C18-C20) within the previous five years. We used multivariable logistic regression analysis to identify factors associated with ICU outcome. Results: During the study period 3650 patients with colorectal cancer were admitted to ICU. Thirteen percent had two or more admissions. The majority of patients (93%) were admitted from surgical specialties with 78% admitted immediately post-operatively. Of those who had undergone a surgical procedure, 34% were performed as an emergency. Median time from diagnosis to ICU admission was 43 (IQR 14-104) days. Median age 71 (IQR 64-78) years; APACHE II score 13 (IQR 0-19); 59% were men. Duke's tumour staging was 12% A, 34% B, 32% C and 10% D with 12% unknown. Organ support was provided by invasive mechanical ventilation (51% of patients), vasoactive drug therapy (51%), and renal replacement therapy (9%). The proportion of patients receiving no, one, two, or three organ support were 35%, 26%, 31%, and 8%. Mortality in ICU was 14.8% (13.0 to 16.7%) among emergency patients, 3.1% (2.3 to 4.0%) among elective patients and 27.1% (24.2 to 30.0%) among non-surgical patients. ICU mortality among Dukes stage A was 9.1% (6.7 to 11.9%), B 11.1% (9.5 to 12.8%), C 11.4% (9.7 to 13.2%), and D 13.9% (10.7 to 17.5%). Mortality was 21.6% (18.0 to 25.5%) among patients with unknown stage. ICU mortality by the number of organs supported was 0.9% (0.1 to 3.2%) for no organ support, 4.8% (2.1 to 9.2%) for one organ support, 20.3% (14.9 to 26.6%) for two organ support, and 39.6% (25.8 to 54.7%) for three organ support. Age, severity score, tumor stage, emergency surgery, being a non-surgical patient and organ support were each independently associated with mortality during ICU when modeled by multivariable logistic regression. Conclusions: Mortality among colorectal cancer patients admitted to ICU was most strongly associated with severity of illness, admitting specialty (surgical or medical), nature of surgery, and number of organs supported during the stay. Tumor stage was independently associated with mortality during ICU stay but showed a smaller differential across stage categories compared to other patient characteristics.