BACKGROUND: The relationship between therapeutic delay and long-term survival from colorectal cancer is unclear. This association was examined prospectively among patients with colorectal cancer in Denmark. METHODS: A total of 740 patients with colorectal cancer were included in a prospective, population-based study in three Danish counties from 1 January 2001 to 31 July 2002. Delay was determined by self-report during a standardized interview. Cox proportional hazards regression was used to compute the hazard ratio (HR) associated with delay, while adjusting for age, sex and co-morbidity, and also for urgency of surgery in patients with colonic cancer. RESULTS: For rectal cancer only, a time span of at least 60 days from the onset of symptoms until treatment (total therapeutic delay) was associated with a 69 per cent higher risk of mortality compared with a total therapeutic delay of less than 60 days (HR 1.69 (95 per cent confidence interval 1.01 to 2.83)). Provider delay (interval from first physician contact until treatment) and hospital delay (interval from referral to a hospital until treatment) of at least 60 days had no impact on survival from colorectal cancer. CONCLUSION: A total therapeutic delay of at least 60 days was a negative prognostic factor for long-term survival from rectal cancer.