The recovery of Pseudomonas aeruginosa on several selective culture media was tested using raw sewage and secondary sewage effluent samples as well as spiked chlorinated imitation swimming water and samples from whirlpools. mPA-medium B gave good recovery of both vital and chlorine-injured P. aeruginosa and selectivity was greater than 90% when analysing whirlpool samples. It is therefore the medium recommended for examination of chlorinated swimming pools. When analysing sewage polluted water with the mPA-B medium, reduced selectivity was noted from low verification rates and from overgrowth by competitive flora. A modified medium (mPA-D; addition of cetrimide, omission of sulphapyridine and actidione) was more selective and sufficiently recovered noninjured cells. Chlorine-injured cells were completely inhibited, however. C-390 (9-chloro-9-(4-diethylaminophenyl)-10-phenylacridan) was confirmed to be highly selective for P. aeruginosa when used in spread plates at a concentration of 30 micrograms/mL; P. aeruginosa was slightly inhibited. However, the medium could not be used with conventional membrane filtration techniques, because cellulose ester filters interfered with the selective action of C-390. Selectivity could be improved by using Gelman Tuffryn (polysulphone) filters and increasing the C-390 concentration to 120 micrograms/mL. At this concentration, however, the medium was strongly inhibitory to P. aeruginosa; resuscitation only partially improved recovery. Two other membrane filtration media were tested. Both cetrimide - nalidixic acid agar and Drake's medium No. 19 were inhibitory to chlorine-injured cells. Several types of membrane filters were tested and there was little difference between them.(ABSTRACT TRUNCATED AT 250 WORDS)