We have previously reported pilot data regarding the safety of saving partially used syringes of a glutaraldehyde cross-linked collagen for use in subsequent treatment sessions with the same individual. That single institution study involved 56 partially used syringes cultured for aerobic bacteria. Only one weakly positive culture was detected among these 56 samples, which prompted us to carry out this expanded study involving multiple centers and different injection techniques. Samples were collected from four centers. Following periurethral injection in an office setting, 166 partially used syringes of glutaraldehyde cross-linked collagen were refrigerated for between 1 and 104 weeks (average 58). Material from all 166 syringes was then cultured qualitatively and quantitatively for both aerobic and anaerobic organisms. Collagen from one syringe grew >100,000 colonies of Escherichia coli. All other cultures were negative. In the pilot study, one culture of 56 syringes was weakly positive for coagulase-negative staphylococcus. When the results from both studies were considered together, only two of 222 partially used syringes (0.9%) were contaminated. The background risk of local infection associated with periurethral collagen injection is approximately 0.29%. Using the statistical equation 'number needed to harm', we found that a clinician would have to reuse 111 syringes at a saving of $34,965 before he or she would cause a single local injection by so doing. Therefore, we feel that it may be cost-effective and safe to reinject material from a partially used syringe of glutaraldehyde cross-linked collagen during a subsequent treatment session on an individual.