The preceding test results appear to favour the impression method used for the testing of surface disinfection procedures. Further investigations were concerned with the selection of a suitable germ carrier. The high recovery rate of the bacteria and the invariably uniform quality of the germ carrier are the decisive criteria. Since a dependence on the type of germ was also to be expected, we checked Staph. aureus, Str. faecalis, E. coli, Kl. pneumoniae and P. aeruginosa for their behaviour on PVC, varnished wooden surfaces, V 2A, rough and smooth glass, resopal, ceramic tiles with dull and glazing finish as well as on plywood. The germ carriers were contaminated quantitatively, dried, sprayed with distilled water to stimulate disinfection and examined after they were allowed to react for a period of 6 hours. A set of 10 germ carriers was examined simultaneously with 10 Rodac impressions. The data were checked by variance analytical methods and evaluated on the mean (or median) values. It was found that while the initial germ number was kept equal OP tiles and V 2A steel produced the highest germ yield. Plywood proved largely unsuitable. Among the test germs, E. coli and Kl. pneumoniae had the lowest adhesion; it was inferior to that of staphylococci and pseudomonads. However, for these 4 germ types 3 impressions per germ carrier suffice for a reliable assessment based on computations. The yield amounts to more than 90 per cent. 94 per cent of the Streptococci were also recovered with 3 impressions, but the absolute germ quantity was low. The use of this type of germ can be recommended only with reservation.