High-level disinfection of GI endoscopes can be reliably obtained under controlled conditions with approved reprocessing methods. However, there are scant data regarding the effectiveness of these methods in clinical practice and no published methods of verification. The purpose of this study is to review retrospectively the results of environmental cultures of flexible endoscopes and to analyze the pattern of results. Cultures of selected GI endoscopes listed as ready to use were obtained by adding 5-15 ml of trypticase soy broth or saline or 30-50 ml of sterile water to the biopsy channel of an endoscope. This wash was collected in a sterile container, plated onto blood and MacConkey agar, incubated at 37 degrees C, and examined for growth at 24 and 48 h. Personnel trained in accordance with approved procedures performed endoscope reprocessing. A total of 312 surveillance cultures were performed between 1990 and 1999. Initially, three of 17 water bottles were found to be contaminated with Pseudomonas species. The bottles were sterilized daily; only sterile water was used and subsequent cultures were negative. Between 1992 and 1994, 15/129 (11.6%) cultures were positive; 14 (93%) were from duodenoscopes. From 1995 to 1997, 18/124 (14.5%) cultures were positive, but only six (33%) were from duodenoscopes. However, 10 (55.6%) positive cultures were obtained from therapeutic upper endoscopes, attributed to faulty mechanical cleaning by nonnursing personnel after emergent procedures. The reprocessing procedure was altered, with improvement. One duodenoscope was persistently culture positive and was found to have a damaged biopsy channel. There were no recognized iatrogenic infections associated with endoscopic procedures. Organisms cultured were commonly gram-negative rods. The use of environmental endoscope culturing is a rapid, simple, inexpensive method to monitor effectiveness of standard reprocessing procedures. Disinfection is less effective with poor mechanical cleansing, and high-titer positivity is a marker for poor cleaning technique. Standard upper and lower scopes are commonly culture negative. Duodenoscopes, because of their inherent complexity, and other scopes used in emergent conditions require particular attention. Surveillance culture results can be used to identify patterns of poor technique, to reinforce proper procedure, and to modify clinical practice. No associated clinical illness was apparent during this study.