In a third international co-operative test of the IBRG Plastics Project Group the validity of a bacterial test was checked in order to improve existing test standards. Nine European laboratories participated in the tests. The test method used was based on DIN 53 739 and as in the preceding two co-operative tests two types of polyurethanes (polyester and polyether) which were differently susceptible to biological attack were employed. For comparative purposes a plasticised PVC used already in earlier tests was also examined. After a test period of 28 days the test specimens were visually evaluated according to a specified numerical evaluation scheme and their weight losses were determined. The results of the visual evaluations varied greatly between the laboratories. Those laboratories that are well experienced with testing succeeded, however, in differentiating between the plastics with the replicates being very uniformly evaluated. The results of the weight loss determinations are more unequivocal. The differences between the biological susceptibilities of the polyether and polyester urethanes were more marked and a statistical analysis of the weight losses of PVC was significant (p = 0.01). As the earlier co-operative tests with mould fungi had already shown (1) a visual examination of the specimens was not sufficient to distinguish clearly between differently susceptible polyurethanes with regard to bacterial attack. (2) Weight loss determinations after a bacterial test are suitable to distinguish between differently susceptible plastics and (3) a test period of 28 days is sufficient.