Aim: The aim was to investigate whether the previously reported causal treatment effect of complete mesocolic excision on the risk of recurrence was biased by inclusion of patients with potentially undiagnosed disseminated disease at the time of surgery, by non-specialist surgery, or caused by mesocolic plane dissection. Method: A population of 1069 patients, 813 undergoing conventional resection and 256 complete mesocolic excision for colon cancer during the period 2008–2013, was stepwise reduced in the following order by excluding patients with recurrence diagnosed within 6 months of the resection, having surgery performed by a non-specialist without supervision, and specimens assessed as not being mesocolic plane dissection. The primary outcome measure was risk of recurrence after 5.2 years using competing risk analyses. Results: The absolute risk reduction of complete mesocolic excision was 6.0% (95% CI 1.8–10.2; P = 0.0049) after excluding patients with recurrence within 6 months of resection, 6.1% (95% CI 1.9–10.4; P = 0.0045) after excluding non-specialist surgery, and 7.5% (95% CI 2.9–12.0; P = 0.0013) after the exclusion of patients whose specimens were assessed as dissections not being performed in the mesocolic plane. Conclusion: The absolute risk reduction of recurrence after complete mesocolic excision for right-sided colon cancer in our previous study was not biased by potentially undiagnosed disseminated disease at the time of surgery or non-specialist surgery, and was not solely caused by dissection in the mesocolic plane. Central vascular dissection with central lymphadenectomy seems a major factor for better oncological results.