This study was designed to examine the current status of - and identify potential gaps in - colorectal cancer outcomes research. English-language articles published from 1999 to 2003 were retrieved from MEDLINE using the search term "colorectal neoplasms OR colorectal cancer" and specific MeSH terms dealing with the effectiveness of interventions on final outcomes and factors, such as costs, involved in those interventions. Articles excluded were biological or pathological studies; studies in non-human subjects; clinical trials or reviews mainly answering efficacy-related questions; and letters, comments, and editorials. Abstracts were reviewed and classified by study design, endpoints, and impact. Full-text articles were retrieved as needed for classification. Of the 2,270 articles retrieved, 525 met the inclusion criteria for the analysis. Most were retrospective cohort studies (216 articles, 41.1%). The most common endpoints were survival (371 articles, 70.7%), followed by quality of life (102 articles, 19.4%), economic cost (81 articles, 15.4%), and patient satisfaction (17 articles, 3.2%). In terms of their impact, 481 articles (91.6%) were classified as level 1 (adds to the knowledge base only); 26 (5.0%); as level 2 (affects practice policies); 16 (3.0%), as level 3 (influences healthcare delivery); and 2 (0.4%), as level 4 (leads to changes in health outcomes). In conclusion, most colorectal cancer outcomes studies published in the past 5 years focused on survival, and the majority represented impact level 1 research. Patient satisfaction was almost ignored. Based on these findings, colorectal cancer outcomes research that answers more than survival questions and represents higher impact levels is warranted.