The number of lymph nodes retrieved from the specimen may be a surrogate measure of the adequacy of extensive colon cancer surgery, but many variables may influence the total lymph node yield of any specimen. We examined which variables would be influential both for negative and positive node sampling.The combined results from 428 patients from three hospitals A to C treated in 2007–2009 with single colon cancers having R0 segmental resections were analysed. The surgical technique and pathology staining methods were slightly different between the hospitals.The mean number of lymph nodes was 15.8 (range 1–60). Twelve or more lymph nodes were harvested in 78% of the specimens. In the multivariate Poisson regression analysis of all TNM stages, the factors associated with the total lymph node harvest were age, pathology handling, tumour location and size (p < 0.001), whereas for TNM stage III alone the pathology handling (p < 0.001) and a radical operating technique (p = 0.003) were highly significant. The total number of lymph nodes was the only significant factor for the number of positive lymph nodes (Posln) according to the multivariate negative regression analysis (p = 0.02) but the analysis of the lymph node ratio (LNR) detected no statistically significant variable.Several factors, and especially the specimen processing technique, were important for the total number of harvested lymph nodes. The number of Posln varied between segments and increased with the total number of harvested lymph nodes, but for LNR no variable was important. LNR seemed to abolish the combined effect of tumour location and the total lymph node yield in prognosis assessment.