Background: Computed tomography-derived body composition parameters are emerging prognostic factors in colorectal cancer. This study aimed to determine the roles of sarcopenia, myosteatosis and adiposity as independent and overlapping parameters in resectable colorectal cancer patients. Methods: This was a retrospective cohort study from a prospectively collected database. It included all adult patients with early-stage colorectal cancer, who underwent curative resection from 2007-09. All patients were seen in a tertiary care cancer center in Northern Alberta. Computed tomography-derived quantification of skeletal muscle and adipose tissues were used to determine cohort-specific cut-offs for sarcopenia, myosteatosis and total adiposity using an optimal stratification analysis. Multivariate cox proportional hazards models were performed to assess for associations between body composition parameters and overall, disease-free and cancer-specific survival. Results: In the 968 patients included, there were a total of 254 disease recurrences and 350 deaths. Body mass index and computed tomography-derived measures of adiposity did not result in worse survival outcomes. Sarcopenia was independently predictive of worse overall (HR 1.45; 95% CI 1.16, 1.84), recurrence-free (HR 1.32, 95% CI 1.00, 1.75) and cancer-specific survival (HR 1.46, 95% CI 1.09, 1.94) in a multivariate model. Myosteatosis was also independently predictive of overall survival (HR 1.53, 95% CI 1.19, 1.97). In a multivariate model considering joint effects of sarcopenia and myosteatosis, the presence of both parameters predicted the worst overall (HR 2.23, 95% CI 1.62, 3.06), recurrence-free (HR 1.53, 95% CI 1.06, 2.21) and cancer-specific survival (HR 2.40, 95% CI 1.69, 3.42). Conclusions: Sarcopenia and myosteatosis are independent predictors of worse survival in early-stage colorectal cancer. Their joint effect is highly predictive of reduced overall, recurrence-free and cancer specific survival. Sarcopenia and myosteatosis represent prognostic factors that may be easily integrated into clinical practice.