Abstract Aims To assess whether active smoking compromises survival in patients with colorectal cancer. Materials and methods We studied a regionally based cohort of 284 consecutive patients referred to the Tayside Cancer Centre for consideration of adjuvant treatment after curative surgery for colorectal cancer. Results Cause-specific survival was significantly worse ( P = 0.0015) in patients who were actively smoking at the time of their first post-operative visit. The absolute difference in 5-year cause-specific survival (active smokers vs the rest) was 21%. In adjusted multi-variate analysis of patients after pathologically complete (R0) resection, the hazard ratio was 2.55 (95% confidence interval 1.40–4.64) in active smokers compared with non-smokers. T stage, number of positive nodes and co-morbidity score were also of independent prognostic influence. Conclusions Persistent smoking was, in this small series, an important and independent predictor of cancer-related death after surgery for cancer of the large bowel. Because smoking and deprivation are related, some of the adverse effects of deprivation upon survival in this group of patients may be explained by smoking behaviour.