OBJECTIVE: To understand disease burden, prognosis, and survival, and estimate target patient populations for treatment, knowing the prevalence of patients by their clinical stage of disease (AJCC/UICC/TNM) is important. However, such data is not readily available in the published SEER (Surveillance, Epidemiology and End Results) statistics. Moreover, the SEER definitions for staging cancer into local, regional, and distant has changed from 1973-98, and do not coincide with the AJCC clinical stages of I, II, III, and IV. The latter has also undergone 5 editions from 1978-1997. Colorectal cancer is the third most common incident cancer in the US, with good prognosis and survival if detected early. We estimated the prevalence and survival of colorectal cancer by clinical stage. METHODS: From the August 2001 release of the SEER database, there were 281,940 cases with a diagnosis of colorectal cancer and complete follow-up through December 31, 1998. The different staging systems of AJCC, TNM, SEER 2000, SEER 1977, SEER Historic Coding , and Dukes/Astler-Collins were summarized and compared, showing the differences and the overlap in the staging systems. Colorectal cancer cases diagnosed from 1988-1998 (n=129,664) were recoded to the AJCC coding system by SEER. In order to estimate age-adjusted prevalence by stage for the year 2001, we calculated the 1998 age specific prevalence. This latter figure was standardized to the 1990 US standard population to estimate the age-adjusted prevalence, which was then projected onto the 2001 US population estimates to calculate an estimated age-adjusted prevalence in 2001 by stage of disease. RESULTS: The estimated year 2001 prevalence of colorectal cancer and median survival by AJCC stage were: Stage I, n=228,958, 9.7 yrs.; Stage II, n=223,936, 6.7 yrs.; Stage III, n=145,307, 3.9 yrs.; and Stage IV, n=28,009, 0.8 yrs. CONCLUSION: Colorectal cancer in the US was found to have decreasing prevalence and survival with increasing clinical stage (severity).