Pelvic abscesses represent the most frequent complications of colorectal surgery. Percutaneous CT or US guided drainage can be an alternative to surgical drainage that is associated to a significant mortality rate. In the current study results of PAD, performed in 39 patients with pelvic or abdominopelvic abscesses were reviewed in order to evaluate reliability of such procedure. Major part of the collections 33/39 (85%) developed after resective colorectal surgery, and 20/39 (51%) were associated to anastomotic fistula; 22/39 (56%) were poorly defined; 16/39 (41%) were multiloculated; 16/39 (41%) had a stool contamination, 23/39 (58%) were greater than 10 cm; 14/39 (35%) were multiple. Thirty-five patients (89.7%) healed, despite high number of complex abscesses and complete resolution of sepsis was achieved in 5.1 +/- 2.9 days. CT proved to be the most reliable tool in assessing distinctive features of collections as well as in identifying the best route for drainage. Adequate size of the catheter was essential to get an effective drainage. In particular, large sized catheter (> 20 Fr) had to be used to drain collections associated to anastomotic fistulas with stool contamination. In four elderly neoplastic patients with chronic illnesses (10%), only a single small sized catheter could be positioned, because of patients poor compliance and PAD was inaffective. Nevertheless even those patients got a partial resolution of the sepsis and their general conditions markedly improved, so that they were able to underwent successful surgical drainage. In conclusion PAD is a safe and reliable tool that can be employed as an alternative to surgical drainage at least as first measure, even if complex pelvic abscesses are found.