used to collect and quantify bacteria from the surfaces of hospital bed mattresses of a single community hospital. Prior to use, the 30 cm2 of the medium on each film plate was hydrated with 0.9 milliliters of buffered peptone water broth. Mattresses were sampled either by rubbing 30 cm2 of mattress surface with amoistened swab or by pressing (10 seconds) the film plates onto areas adjacent to the areas sampled by the swabs. Swabs were applied to 5% sheep blood agar medium (BA) and streaked for isolation; growth was examined after BA was incubated 18-24 hours at 378C in 5-7% CO2. Film plates were incubated as above and then blotted onto BA, which was incubated as above. The McNemar test was used to compare the results of the cultures. Results: 91 mattresses were sampled by both methods: 62 from mattresses that had undergone terminal cleaning, and 29 from mattresses that were dirty and ready for terminal cleaning. Comparing samples from the same mattress: 34 (37.4%) of the mattresses produced growth by both methods; 6 (6.6%) produced growth by neither method, 46 (50.5%) produced growth by the film plate method but no growth was identified by the swab method; 5 (5.5%) produced growth by the swab method only. The difference between the sensitivity of the two methods was statistically significant (p,0.001). Bacteria isolated by both methods represented normal skin bacteria and environmental bacteria expected in hospital patient rooms; the isolates included; Staphylococcus species, Bacillus species, Gram-negative rods, Micrococcus species, Streptococcus species, and diphtheroids.