Background The Physiological and Operative Severity Score for the enUmeration of Mortality andmorbidity (POSSUM) and later modifications (P-POSSUM y CRPOSSUM) have been used to predictmorbidity and mortality rates among patients with rectal cancer undergoing surgery. These calculations needsome adjustment, however. The aim of this study was to assess the applicability of POSSUM to a group ofpatients with rectal cancer undergoing surgery, analysing surgical morbidity by means of several variables. Methods between January 1995 and December 2004, 273 consecutive patients underwent surgery forrectal cancer. Information was gathered about the patients, tumour and therapy. To assess the predictioncapacity of POSSUM, subgroups for analysis were created according to variables related to operativemorbidity and mortality. Results The global morbidity rate was 23.6% (31.2% predicted by POSSUM). The mortality rate was 0.7% (6.64, 1.95 and 2.08 predicted by POSSUM, P-POSSUM and CR-POSSUM respectively). POSSUMpredictions may be more accurate for patients younger than 51 years, older than 70 years, with low anaesthetic risk (ASA I/II), DUKES stage C and D, surgery duration of less than 180 minutes and for thosereceiving neoadjuvant therapy. Conclusion POSSUM is a good instrument to make results between different institutions and publicationcomparable. We found prediction errors for some variables related to morbidity. Modifications of surgicalvariables and specifications for neoadjuvant therapy as well as physiological variables including life stylemay improve future prediction of surgical risk. More research is needed to identify further potential riskfactors for surgical complications.